Full text of "Acta Physiologica Scandinavica"

See other formats

ACTA

PHYSIOLOGICA

SCANDINAVICA

VOL. 1

REDACTORES

F. Reenpaa a. Krogh E. Langfeldt

Helsinki Kjohenhavn Oslo

G. LiLJESTRAND (EDITOR)

Stockholm

COLLABORANTES

G. AitLGREN (Lund), Y. Airila (Helsinki), E. L. Baoksian (Uppsala),

G. Blix (Uppsala), J. Bock (Kjobenhavn), R. Ege (Kjobenhavn),

H. V. Egler (Stockholm), U. S. v. Euler (Stockholm), A. Folling
(Oslo), R. Granit (Stockholm), G. Gothlin (Uppsala), E. Hammarsten
(Stockholm), E. Haksen (Kjobenhavn), K. Hansen (Oslo), G. Kahlson
(Lund), F. Leegaard (Oslo), J. Lehmann (Goteborg), J. Lindhard (Kjo-
benhavn), E. Lundsgaard (Kjobenhavn), R. Nicolaysen (Oslo), S. Or-
SKOv (Aarhus), A. V. Sahlstedt (Stockholm), F. Sohonheyder (Aarhus),
P. E. SiMOLA (Helsinki), T. Teorell (Uppsala), H.Theorell (Stockholm),
T. Tuunberg (Lund), A. Westerlund (Uppsala), E. Wibsiark (Lund),

A. I. ViRTANEN (Helsinki)

1940-1941

Reprinted with the permission of Acta Physiologica Scandinavica
KaroUnska Institiirer, Stockholm

JOHNSON REPRINT CORPORATION Johnson Reprint Company Limited
111 Fifth Avenue, New York, N.Y. 10003 Berkeley Square House, London, W.l

First reprinting, 1964, Johnson Reprint Corporation

Printed in the United States of America

YOL. I. INDEX.

Ease. 1. (30. IX. 1940.)

Pag-

Introduction 1

A ^Icthod of Blood Volume Determination. By L. Hahn and G.

Hevksy 3

Rate of Pa.ssage of Water through Capillarj’^ and Cell Walls. By

G. Hevesy and C. F. Jacobsen 11

Zur Pharmakologie der Gozymase. Von H. v. Euler und U. S.

V. Euler 19

Studies in Glycoproteins. By Gunnar Blix 29

A Colorimetric Method for the Determination of Potassium in

O.oi — 0.1 cc of Blood Serum. By K. A. J. Wretlind 43

A New Slethod for the Determination of Carbon Monoxide in

Blood. By R. Wennesland . .' 49

A Parapharyngeal Jlethod of Hypophysectomy in Rabbits. By

Dora Jacobsohn and Axel Westman 71

SulfapjTidinc in Secretin Stimulated Pancreatic Juice and Bile

of Cats. By A. Taylor and G. Agren 79

Some Quantitative Data on the Antagonism between Piperido-
Methyl-Benzo-Dioxane (933 F) and Adrenaline. By N.-O.

Abdon and S. 0. Hammarskjold 85

Wirkung der Sauerstoffatmung auf die Atmungssteigerung bei Ca-

rotisabkleramung. Von Tube Rudberg 89

The Effect of Carotid Sinus Denervation on Respiration during
Rest. By U. S. v. Euler and G. Liljestrand 93

Ease. 2. (15. XL 1940.)

The Interaction of Amino Nitrogen and Carbohydrates. By Gun-
nar Agren 105

On the Peptidase Activity in the Cattle Muscle. By Gunnar

Agren 119

Die Stabilitat der organischen Phosphorverbindungen und Phos-
phatase in Pferdeblut bei dessen Aufbewahrung in vitro. Von
Knot Sjoberg 125

VOI,. 1. INDKX.

Studies on the Muscular IMiysiology of the Genital Traci. I. The

Spontaneous Tonus of the Uterine Muscle; Its Deiiendcnce on

Hormonic Factor.s. liy Sunk Genkul 130

Method for Rapid Determination of Specific (Jravity. liy C. F.

Jacobsen and K. LiNnEP..STiH5M-LANO 110

On the Influence of Atropine on .some Xicotinelike Aefion.s of

Acetylcholine. By X.-O. Abdon loJ

Rate of Penetration of Phosphate into Muscle Cell.':. By G, He-

VESY and 0. Rebbe 171

Uber den Einfluss lokalcr phy.::ikali.scher und chenii.«chcr ITaut-
reize auf die periphcrc Blutvertcilung. Yon Gosta vo.n Rees
und Fritiof SJosTitANU 18.3

Fasc. 3. (iss.xir. iti io.)

Tension Changes during Tetanus in Mammalian and Avian Muscle.

By U. S. VON Euler and Roy L. Swank 203

Pyrophosphatase im Bitit. Von Knut S-Iobero 220

Transamination with Pc])tide Sul>.stratc.s in Cattle Diaphragm

Muscle. By Gunnar Aoren 233

The Absorption of Ethyl Alcohol from the Gastro-Intcstinal Tract
as a Diffusion Process. By Sven M. liEROc.RKN and Ekonaru

Goldberg 240

Evidence for two Phase.s in the Regeneration of Visual Purple.

By M. Zewi ' 271

Eine einfache und fiir klinische Zwceke geeignetc 31ikromethode
zur Bestimmung des ITarnstoffstickstoffs (ilrX) im Blutc durch
Ureasebehandlung und diroktc Xcsslcrisicrung. Von Wilhelm

Ohlsson 278

In vitro Studies on the Role of Vitamin D in the Metabolism of
Calcium and Phosphorus in the Rat Bones. By V. V. Krae-
MER, B. Landtman and P. E. Simola 285

Fasc. i, (20. II. 1041.)

Uber den Einfluss lokaler Haut-schadigungen (mechanische Ver-
letzung, Erfrierung, Verbrennung) auf die ])eriphere Blutver-

teilung. Von Go.sta von Reis und Fritiof Sjostrand 299

Electrothyreogram, Blood Iodine and Thyroid Iodine during Sti-
mulation of the Sympathetic. By B. Heltn and H. Zilliacus 317
Uber die Wirkung von Digitalis, Cardiazol, Coramin, Hexeton
und Strychnin auf Kreislauf und Atmung des gesunden Men-

schen. Von 6. Liljestrand und G. Xyltn 328

Die Wirkung des Atropins auf das Hcrzminutenvolumcn des ge-
sunden Menschen. Von R. Domenjoz 339

Rate of Penetration of Ions through the Capillary Wall. By' L.
Hahn and G. Hevesy

347

VOL. 1. INDEX.

Pag.

Studies on the Muscular Physiology of the Genital Tract. II.
Tonus, Spontaneous Activity and Drug Reactions in the Cervical

Mnsclcs. By Sune Genell 362

Rotation of Activity and Spontaneous Rhythms in the Retina.

By Ragnar Granit 370

Dark-Adaptation and the Platinum Chloride Methode of Staining

Visual Purple. By Sten Stenius 380

t)ber den Reizmechanismus dcr Chcmorczeptoren im Glomus
caroticum. Von U. S. v. Euler, G. Liljestrand und Y. Zotter-

MAN ^ 383

The »Red» Receptor of Testudo. By Ragnar -Granit 386

An Attempt at Isolating the Carbohydrate Moiety of Crystalline
Ovalbumin by IMeans of Cataphoresis. B}' Erik Jorpes and
Torsten Tiianing 389

Optic Pathway. By Carl Gustaf Bernhard.
ipplementum II. On the Citric Acid Metaboli
13y Johan MArtensson.

VIII

INDEX ATICTOUUM.

Pag.

ZoTTERMAN, Y., U. S. V. Euder find G. Liwestrand, Reizrnecha-

iiismu.B der Chemorczcptorcn 383

Agren, G., Amino Nitrogen and Carbohydrates 105

Agren, G., Peptidase Activity in Cattle Muscle 119

Agren, G., Transamination in Cattle Diaphragm 233

Agren, G., and A. Taylor, Sulfap}'ridinc in Pancreatic Juice
and Bile ; 79

Errata.

P. 109, line 1, from above, for “did not accelcrnte” read “accelerated".

P. 116, line 7, from above, for “dit not facilitate” read “facilitated”,

P. 116, line 2, from below, the formula to the left read;

'''KH, +

i i

II, • c — ccoon

INTRODUCTION.

The intimate cultural and personal relations between the Scan-
dina\’ian countries, Denmark, Finland, Norway and Sweden,
already long ago led to the foundation of common scientific jour-
nals. In this way the contributions from these countries were to a
great extent made easily accessible to the Scandinavian scientists
themselves, and collaboration and mutual understanding were
facilitated. Later on, such journals were made generally available
by being published in English, French or German, thus giving an
opportunity to scientists from all parts of the world to become
acquainted with the work performed in Scandinavia. There can
be no doubt that this has led to an increased knowledge and appre-
ciation of what has been done and stimulated to new efforts.

Witliin the medical sciences the first journal of the type men-
tioned was started in 18G9 under the title Nordiskt Medicinskt
Arkiv, It became necessary, however, to divide it into two ar-
chives, one for medicine and one for surpry. From 1919 these
have appeared as Acta medica scandinavica and Acta chirurgica
scandinavica. A large number of other Acta has been added to
those already existing, there being at present 11 such journals,
covering different fields within the medical sciences.

With regard to physiology and related subjects it was found at
an early date that a special journal was highly desirable. Several
difficulties made it impossible, however, to start suck an archive
entirely within Scandinavia. A practical solution was found,
when the German firm Veit & Co. in Leipzic in 1889 undertook
to publish the Skandinavisches Aichiv fur Physiologie. For more

1 — i01S23. Acfa phvs. Scandinav. Tol.l.

IKTUODUCTION,

than 50 years this journal has existed. In all 83 volumes and 10
supplements have appeared. They contain 1 308 contributions
from 684 different authors. Only in exceptional cases have papers
from Non-Scandinavians been accepted — mostly in connection
■with special jubilee volumes. Of the Scandinavian authors 118
are from Denmark, 191 from Finland, 22 from Norway and 276
from Sweden. The Scandinavdan physiological Society, founded in
1926, has published an account of the proceedings of her congres-
ses in the journal. Undoubtedly it 1ms been of great value for
Scandinavian physiology for half a century, and we owe the Ger-
man firms Veit k Co. and Walter de Gruyter our sincere thanks
for their valuable assistance.

The old idea of a physiological journal, owned by the Scandina-
vian physiologists and printed in Scandinavia, has never died,
and the question has been taken up and discussed at numerous
occasions. Recently a decision has been taken that the Skandina-
visches Archiv shall be discontinued and a new journal started
under the name of Acta phpiologica scandinavica, belonging to
the Scandinavian physiological Society and printed in one of the
Scandinavian countries. A number of representatives of the ph 5 's-
iological sciences in these countries have promised their colla-
boration. The following will act as editors: for Denmark Profes-
sor A. Kbogh, for Finland Professor Y. Reextaa, for Nor^vay
Professor E. Langfeldt and for Sweden Professor G. Lieje-
STBAND. The journal will contain contributions to Physiology,
Biochemistry and Pharmacology by ScandinaNuan authors or
from Scandina'vian laboratories. The articles will be published
in English, French or German. Not more than two volumes ■inll
appear each year. In special cases supplements will appear; they
will be sent to the subscribers without extra cost.

We hope that Acta physiologica scandinavica will promote
physiology and meet with the approval of our colleagues through-
out the world.

G. Liljestrakd.

From the Institute of Theoretical Physics, Copenhagen.

A Method of Blood Tolume Beteriiiiiiation.*

L. HAHN and G. HEVESY.

(With 1 flgnro in the text.)

The method usually applied in the determination of blood
volume is that vorked out by Rowntkee and his colleagues
(1929). The principle of the method is that a dyestuff is injected
intravenously and its degree of dilution determined.® As the dye
only mixes with plasma, the volume of the plasma alone is thus
measured. The relative volume of corpuscles and plasma is
determined with the haematnerit. To arrive at the blood volume
the volume of the corpuscles is added to that of the plasma.

Rowxtree gives the following description of the method
applied (comp, also Fleischer-Hansen, 1928). A 1.5 per cent
solution of ^’ital red in distilled water is prepared. Four centrifuge
tubes are pro\dded and 1 c.c. of a 1.6 per cent solution of sodium
oxalate is placed into each of them. A needle is inserted in the
vein of one arm and 10 c.c. of blood are removed. 5 c.c. are placed
into each of two centrifuge tubes for standard plasma colour.
The dye is then injected. After 3 to 6 min., 10 c.c. of blood are
withdrawn from the vein of the other arm and 5 c.c. placed into
each of the two remaining centrifuge tubes. All four tubes are
centrifuged and the relative volume of corpuscles and plasma
measured. The second sample is compared with a known strength
of the dye and the degree of dilution of the dye in the plasma is
thus obtained.

‘ Eeceived for publication 29 March 1940.

* Instead of a dyestuff, diphteria antitoxin "was used in some determinations
(V. Beiibing, 1912; JIadsen, 1934).

L. IIAIIN AND O, IIEVESY,

When considering the possible errors of this method, the main
question at issue is whether, when the second sample is collected,
the dye is uniformly mixed in the plasma and none has yet escaped
into the tissue spaces or urine, a further possible source of error
being the adsorption of a part of the dye by the enormous surface
of the capillary wall.

Determination of blood voinme based on tlie dilution
of labelled corpuscles.

In this note, we msh to describe a method of blood volume
determination based on an entirely different principle from that
described above. We inject into the vein of a rabbit A a known
volume of labelled corpuscles taken from another rabbit B and
determine the extent to which these labelled corpuscles arc diluted
in the circulation of rabbit A. Labelled corpuscles of rabbit B
are obtained in the follo^ving way. We administer by subcutaneous
injection some labelled (radioactive) sodium phosphate to rabbit
B. In the course of about a week, a substantial fraction of the
phosphatide molecxdes of the bone marrow and other organs are
renewed. As this renewal takes place in the presence of labelled
phosphate, the newly formed phosphatide molecules will contain
labelled P atoms. Corpuscles formed in a medium containing la-
belled phosphatide molecules will necessarily incorporate some of
them. Labelled phosphatide molecules can also enter to some
extent into the corpuscles by exchange of non-active phosphatide
molecules with active phosphatide molecules present in the plasma.
The various ways of incorporating labelled phosphatides into
corpuscles are described in detail in a paper which is in print
(Hahn and Hevesy, 1940).

Besides labelled phosphatides, labelled varieties of several acid-
soluble organic phosphorus compounds as, for example, those of
glycerophosphate and adenosincriphosphate, are found in the
corpuscles. Each of these compounds can be used as an indicator
when determining the dilution of the corpuscles of rabbit B in
the circulation of rabbit A. It is, however, more convenient to
extract the total acid-soluble P and to use the mixture obtained
as an indicator.

BLOOD VOLTOIE DETERMINATION.

Determination of the blood Yolinue of a rabbit
'neighing 2 hg.

a) Mating use of the labelled phosphatides of the corpuscles.

IVe admiuistered radioactive sodium pliosphate of negligible
weight ha-ring the actirity of about O.OOi milliCurie to rabbit B.
After the lapse of a week, 1 c.c. of blood of rabbit B containing
0.32 c.c. corpuscles was injected into the jugularis of rabbit A.
After the lapse of 5 min., 50 c.c. blood were collected and, after
the addition of heparin, centrifuged. The haematocrit value of
this sample was found to be 0.33. The phosphatides of the cor-
puscles were thoroughly extracted by Bloor’s method. Their P
was converted by wet ashing into phosphate. The phosphate was
precipitated as ammonium magnesium salt. Before precipitation,
sodium phosphate was added to the solution to obtain a pre-
cipitate of about 80 mg. The actirity of the precipitate was then
determined by means of a Geiger counter. The comparison of
the actirity of the samples is facilitated if they have practically
the same weight and it is for this reason that we added to the
original sample a comparatively large amoxmt of non-active
phosphate. The corpuscles of 1 c.c. of the blood of rabbit B were
also extracted with ether-alcohol and the extract treated in the way
described above. The actirity of the sample thus obtained was
then compared with that of the corresponding sample of rabbit A.

Let us denote the injected blood volume by Vj, the volume of
the sample collected for analysis from rabbit B and rabbit A,
respectively, by Vi and Vjj, and the activity of the two samples
obtained by Aj and An; then the blood volume to be determined
(X) becomes

_ ArVH-V.

An -V i

In some of our experiments, before injecting, for example, 1 c.c.
blood into the jugularis of rabbit A, we removed 1 cc. In that
case, the second term of the equation becomes 0.

Operations involved in the determination of the total blood
volume are thus: measurement of the volume of three samples,
and the comparison of the radioactivity of two samples.

In the above mentioned experiment the corpuscle phosphatides
of 1 c.c. blood of rabbit B contained 100 relative actirity units;

L. HAHN AND G. HEVESY.

th.e activity of the corpuscle phosphatides extracted from 50 c.c.
blood of rabbit A "Was found to be 53.3. From these values it
follows that the blood volume of rabbit B amounts to 93.8 — 1
c.c. = 92.8.

b) Mating use of labelled acid-soluble compoimds
of the corpuscles.

We can check the result obtained above by another procedure
in which, instead of the labelled phosphatides, the labelled acid-
soluble phosphorus compounds are involved. After extraction of
the phosphatides, the corpuscles are extracted with 10 per cent
trichloroacetic acid. The P of the filtrate obtained is converted,
as described above, into ammonium magnesium phosphate. The
activity of the sample obtained from rabbit A is compared with
that of the sample from rabbit B, as described above. The figures
obtained being 100 and 54.4 respectively, the total blood volume
of rabbit A becomes

Ax-Vn’Vi

An-

100*50 -1
54.4*1

1 = 91.9—1 = 90.9.

The use of labelled acid-soluble P compounds leads thus to
practically the same result as obtained with labelled phosphatides
as indicators.

The blood volume per kg of rabbit weight was found, in the
experiment described above, to be 46 c.c. In another experi-
ment the value of 38 c.c. was found.

Determination of the blood volume of a chick
weighing 135 gra.

Labelled sodium phosphate was administered to chick B and,
24 hours later, 0.5 c.c. blood of this chick was injected into the
jugular vein of chick A. The haematocrit values were found
to be 0.26 and 0.28, respectively. We found the activity of
the phosphatides extracted from the corpuscles of 1 c.c. blood
of chick A to be 7.6, taking that of the phosphatides secured from
the corpuscles of 1 c.c. blood of chick B to be 100. The total blood
volume of chick A is thus

X =

100*1*0.5

7.6*1

0.5 = 6.1,

or 45 c.c. per kg weight.

BLOOD VOLUME DETERMINATION.

Loss of labelled P compounds by the corpuscles.

We visli noTT to discuss the possible loss of labelled P compounds
by the corpuscles during the interval between the injection of
labelled corpuscles of rabbit B into the circulation of rabbit A
and the securing of blood samples of rabbit A. Such a loss would
clearly involve a source of error by leading to too high values for
the blood volume to be determined. As to the loss of labelled
phosphatides, we found that the labelled phosphatide content
of the corpuscles of rabbit A, after the lapse of 1 hour, amounts
to 90 per cent of that found after the lapse of 6 min. This result
shows that the disappearance of labelled phosphatides from the
corpuscles takes place at a slow rate.

As to the disappearance of the labelled acid-soluble P molecules
from the corpuscles, we find the following result. The labelled
acid-soluble P content of 1 c.c. corpuscle of rabbit A, 8.5 min.
after injecting the blood of rabbit B into rabbit A, is, as seen
in Table 1 and Fig. 1, within the errors of experiment, identical
with that found after 2.5 min.

Table 1.

Disappearance of labelled acid-soluble P compounds
from ihe corpuscles.

Time elapsed after injecting
the blood of rabbit B into
rabbit A

Per cent of labelled acid-
solnble P injected, present
in 1 cc corpuscle of
rabbit A

2.5 min. •

3.89

5 >

3.21

6.5 J

3.39

8.5 >

3.80

26.5 >

3.12

60 » .

2.88

120 »

2.58

210 »

1.35

We have now to investigate if, during the lapse of 5 min. or
more, after which time blood samples in the experiments described
in this paper were secured, a uniform mixing of the labelled cor-
puscles of rabbit B in the circulation of rabbit A took place. From

lABEUED ACIO-SOlUBLE P CONTENT OF CORPUSCLES

L. HAHN AND G. HBVESY.

Fig. 1. Disappearance of labelled acid-aolnble P from corpuscles, injected into

the circulation of a rabbit.

tlie data contained in Table 1 and Fig. 1 we can conclude that,
already after the lapse of 2.5 min., a uniform mixing actually
took place. Such a result in the case of a small animal with fast
circulation could be expected since, in experiments carried out
on human subjects, a uniform mixing of dyestuffs injected into
the plasma was found to take place in the course of 6 min. (Rown-
TREE et alia, 1929; Grape et alia, 1931).

Discussion.

The sole difference between the normal and the labelled cor-
puscles is that in some of the molecules present in the normal
ones P (having the mass number 31) is replaced by radioactive P
(having the mass number 32). In 1 c.c. corpuscles of rabbit B,
0.06 mg phosphatide P was present. Of these 0.06 mg 10~ii mg
were radioactive ®-P atoms. As 1 c.c. contains about 3 • 10® cor-
puscles, one corpuscle contains on an average 10'®^ mg ®®P or
only about one corpuscle in a himdred contained an active
phosphatide molecule. After injecting the blood sample of rabbit
B into rabbit A, a strong dilution of the labelled phosphatides
took place: only one in about three thousand corpuscles now
contains a radioactive phosphatide P atom. The replacement of
a minute percentage of the ®'P atoms by ®®P atoms in the P com-
pounds of the corpuscles can hardly influence to any noticeable
extent the chemical properties of the corpuscles and we can,
therefore, claim that, when appljdng the method described in this

BLOOD VOLUME DETERMINATrON.

note, no non-physiological component is introduced into tlie cir-
culation. As to the /S-radiation emitted bj the ®-P atoms present
in the corpuscle phosphatides, the number of /?-particles emitted
per minute in the total circulation of rabbit A amounts to only
about 1000, vhile the number emitted by the total acid-soluble
fraction amounts to about 30 times that figure. Hovr insignificant
these figures are can best be realized when v-e envisage that this
radiation corresponds to that of only 10~® and 10"® gm radium,
respectively.

li\Tien carrying out experiments as those decribed above on
human subjects, it may be advisable to make use of the acid-
soluble P compounds of the corpuscles as indicators. Since, in
this case, one may use less radioactive P, such experiments can be
carried out on human subjects by administering by subcutaneous
injection or by mouth to the blood donor sodium phosphate
ha\’ing a /5-radioactivity corresponding to that of about 0.01
milliCurie or even less.

Summary.

A method of blood volume determination based on the deter-
mination of the dilution of labelled corpuscles is described. Radio-
active sodium phosphate is administered to rabbit B; after the
lapse of some days, a known blood sample of this rabbit is in-
jected into the vein of rabbit A. A few minutes later, the corpuscles
of a known volume of the blood of rabbit A are secured, their
phosphatide content extracted, and its activity measured. Cor-
puscles of a known blood volume of rabbit B are treated in the
same way. From the ratio of the labelled phosphatide P content
of the corpuscles of rabbit A and rabbit B, the total blood volume
of rabbit A is calculated.

An alternative and often preferable determination is based on
the comparison of the activity of the acid-soluble P secured from
the corpuscle samples of rabbit A and rabbit B.

The blood volume per kg of rabbit weight is found to be 42 c.c.,
per kg of chick weight 45 c.c.

We wish to express our hearty thanks to Professor FTlels Bohr
for putting numerous facilities at our disposal, to Professor
Ernest 0. Lawrence for the gift of the radioactive phosphate,
and to Dr. H. Dam for his effective help in the experiments with
chicks.

L. HAHN AND Q. HEVESY.

Literature.

V. Behring, E., Beitr. z. exp. Ther. 1912, 12 , 2.

Fleischer-Hansen, C. C., Studier over Blodvolumen, Copenhagen
1928.

Graff, S., 0. A. d’Esopo, and A. J. B. Tillmann, Arch. Intern. Med.
1931, 48 , 821.

Hahn, L. and 6 . Hevesy, Kgl. Danske Vidensk. Sclsk. Bipl. Medd.
1940 (in print).

Madsen, E., Acta path, microbiol. scand. 1934, 11 , 376.

Eowntree, L. 6 ., G. E. Brown and G. M. Roth, The Volume of the
Blood and Plasma in Health and Disease, Philadelphia 1929.

Institute of Theoretical Physics and The Carlsberg Laboratory,

Copenhagen.

Rate of Passage of Water tlirougli Capillary

and Cell Walls.

G. HEVESy and G. F. JACOBSEN.^

(With 1 figure in the text).

Water molecules, tvhicli are absorbed into the circulation, ■vrill
mix rapidly with those present in the plasma. They wdll then
penetrate the capillary wall and become distributed in the extra-
cellular fluid. Ultimately, they will invade the cells. Simultane-
ously, a loss of some of the water molecules through the kidneys,
the bowels, the lungs and tlirough peripheral evaporation will also
take place. It is difficult to estimate even very roughly the rate
at which some of the above processes take place. Experiments in
which heavy water is used as an indicator permit, however, the
determination of the rate at which individual water molecules
introduced into the circiilation are distributed in the body water
and from these determinations to answer the above questions.

We inject a few cc. of practically pure heavy water into the jugu-
laris of a rabbit and take at intervals blood samples from the
carotis. The next step is to prepare pure water from the blood
samples® and to determine its density. Let us assume that we
inject 1 cc. of heav}’- water having a density of 1.1000, and find
for the water prepared from a blood sample the value 1.001.
Then we must conclude that the 1 cc. heavy water injected into
the vein was diluted by 99 cc. of normal water present in the body
of the rabbit in the course of the time which elapsed between the
injection of the heavy water and the collection of the blood sample.

* Received for publication 11 April 1940.

’ The total water content of the blood has to bo distilled to avoid a fractiona-
tion of the diluted heavy water, the vapour pressure of deuterium oxyde being
smaller than that of H.O.

G. IJKVESY AND C, F. JACOBSEN,

Experimental procedure.

Heavy water in contact with air containing vapour of normal
water rapidly becomes lighter in consequence of the interchange in
the vapour phase. In view of this interchange, it was necessary
to keep the samples containing heavy water out of contact for
any prolonged time with the moist atmosplicrc. The blood samples

of about 1 cc, volume were collected in small dishes containing
traces of heparin, the blood was transferred into part A of the
glass vessel, seen in Fig, 1, and the vessel closed with soft parafin.
These operations took only a few seconds. About Yj nig-
P-Oj was added to the solution in order to neutralize any traces
of ammonia present. The vessel was then cooled in the refrigator,
evacuated, and sealed off at B. Tube D was then immersed in
liqrdd air, while the other parts of the vessel remained at room
temperature. After the lapse of a few hours, the water content of
the blood sample was found to be present in the form of ice in
tube C. After the ice was molten and the water collected in D,
this tube was sealed off. The pure water obtained by this proce-

PASSAGE OF WATER THROUGH CAPILLARY AND CELL WALLS. 33

dure was further purified by distillatiou in the presence of potas-
sium permanganate and sodium peroxyde, and the density of the
purified samples determinedj using Linderstrom-Lang’s floating
drop method (Linderstrom-Laitg, Jacobsen and Johansen
1938).

Experiment JL.

In tliis experiment, a rabbit weighing 2.6 kg. was used. 9.2 cc.
of heavy water ha\dng a density of 1.1049 were injected. The time
which elapsed after the injection of the hea\’y water is recorded
in the first column of Table 1, while the next column contains
data on the density excess of the blood water over normal water,
expressed in parts per million. (The heavy water injected had a
density excess of lOI 900 parts per milli on.) The third column
contains data on the dilution of 1 cc. heavy water introduced into
the circulation. In the fourth column the percentage of the weight
of the rabbit which took part in the dilution process is stated.

Table 1.

Time

Density excess of
blood vater in parts
per miUion

Extent of dilution
of 1 cc. heavy
■water

Diluting ■water
volume expressed
in percentage of
body ■weight

40 sec. . . .

1500

644

24.7

185 sec. . . .

1300

742

28.6

8.3 min. . .

930

1038

40.0

6.1 min. . .

795

1221

46.9

24.1 min. . .

527

1825

70.2

5 days . .

339

2 843

109

39 days . .

10810

417

The injection lasted 40 sec.; the time recorded in column 1 is
calculated from the moment half of the water was injected; the
first blood sample was collected 20 sec. after all the heavy water
was injected. In the course of such a short time as 40 sec., 644 cc.
of the body water took part in diluting the heavy water injected
and, after the lapse of 24 min., as much as 1 825 cc.

The plasma water cannot diffuse into the cells without passing
the capillary wall. If the last mentioned process took place within
40 sec., then the volume of the diluting water should be at least
equal to that of the extracellular fluid which amounts, in a rabbit

Q, HEVBSY AND C. I'. JACOBSEN.

weighing 2.G kg, to 670 cch As seen in column 4, the volume of
the diluting water was only slightly less, namely 650 cc,, than this
value. The total water content of the rabbit amounts to 70 to
75 per cent of its weight, corresponding to a volume of 1820 to
1 950 cc. As seen in column 4, the sample collected after the lapse
of 24 min. was found to be diluted by 1825 cc. body water. Within
that time, therefore, a distribution of the heavy water in almost
the total body water took place, though some of the water present
in certain organs may not have taken part in the exchange pro-
cess=. This point can only be settled by investigating the density
of tissue water. The above figures suggest that, in contradistinc-
tion to a very fast invasion of the interspaces, the penetration
into the cells is a somewhat slower process.

In the samples collected shortly after the start of the experiment
the dilution, due to a loss of heavy water by the body, can be dis-
regarded. This is not the case in experiments lasting several hours
or days. In the course of 5 days, for example, the loss of water
through the kidneys alone amounts to about 1 liter, thus to 38 per
cent of the rabbit’s weight. In this case, the excretion of a corres-
ponding part of the heavy water is responsible for the at first
sight puzzling value of 109 per cent found. The hydrogen atoms
bound to oxygen or nitrogen in the various organic compounds
present in the body exchange speedily with those present in the
w'ater or heavy water molecules, and this process will also increase
the dilution figures observed, as a removal of deuterium acts in
the same way on the water density figures as does dilution by nor-
mal water. In view, however, of the fact that the amount of
hydrogen present in the organic compounds is small compared
with that of the hydrogen incorporated in water molecules, the
process mentioned above ^vill not much influence the dilution
figures obtained. In experiments of long duration, a successive
replacement of most the hydrogen atoms present in organic
molecules will take place, giving an additional outlet to some of
the deuterium atoms present in the body water. The percentage
of hydrogen present in the fats of the body which exchanges within
1 hour with water hydrogen is negUgible; the corresponding amount
of protein hydrogen is not (Ussing 1938). The water equivalent
of this hydrogen amounts, however, only to to ^ per cent of
the body weight, or 13 to 52 cc. in the case of rabbit A and 8 to 30 cc.

' Comp., for example, A. Kbooh (1937).
* Comp. UssiKQ (1938).

PASSAGE OF WATER THROUGH CAPILLARY AND CELL WALLS. 15

in the case of rabhit B. The amount of catabolic water formed in
the course of 1 hour in the rabbits amounts only to about 0.1 per
cent of the body weight. As seen in Table 1, after the lapse of 39
days, the density excess of the blood water very much declined.
This decline is mainly due to loss of the heavy water and, thus,
of a corresponding amount of normal water present at the start of
the experiment from the body. About 6/6 of these molecules was
lost in the course of 39 days. In the case of human subjects,
who drank heavy water, it was found (Hevesy and Hofer, 1931)
that, in the course of 9 days, half of the heavy water taken was lost.

In the above connection, it is of interest to recall the experiments
carried out by McDougall, Verzar, Erlenmeyer and Gaert-
NER (1934). They injected a solution of heavy water into the jejun-
al loops of rats and investigated the heavy water content of the
intestinal fluid of the rats killed 1 hour after the start of the experi-
ment. They found the diluting water volume of the rat to amount
to 66 per cent of the body weight.

Experiment B.

This experiment -was carried out on a rabbit weighing 1.5 kg.
5.0 cc. of heavy water were injected. The injection took 6 sec.
The first blood sample was taken in the interval of 22 — 26 sec.
after the start of the experiment. The time recorded is reckoned
from the moment that half of the heavy water was injected until
half of the blood sample was collected.

Tnhlo 2.

Time

Density excess of
blood •water in parts
per million

Extent of dilution
of 1 cc. hea'vy
•water in the circu-
lation

Dilating water
volume expressed in
percentage of body
weight

21 sec. . . .

1034

506

80 sec. . . .

856

612

40.8

1.8 min. . . .

794

661

44.1

3 min. . . .

719

727

47.8

5.2 min. . . .

570

921

• 61.4

7.9 min. . . .

495

1056

70.4

13 min. . . .

490

1070

71.3

22.2 min. . . .

440 ,

1190 ,

79.4

30 min. . . .

450

1167

77.9

48 hours . .

412

1274

G, IIEVESY AND C, F, JACOBSEN,

While rabbit A had not shown any sign of distress after the blood
samples were taken, this was not the case with rabbit B. In the
course of the two days following the start of the experiment, only
a small amount of urine was produced by rabbit B, with the con-
sequence that no pronounced effect due to loss of heavy water
through the kidneys is shown by the density figures of the blood
water sample of rabbit B collected after the lapse of 2 days, in
contradistinction to the results obtained when investigating the
blood water of rabbit A after the lapse of some days.

Discussion.

We found that, within about Ys min., heavy water injected into
the jugularis of rabbits was diluted by a large amount of body
water, the volume of which corresponds to about that of the extra-
cellular space of the body. This rapid dilution is followed by a
second, slower process, presumably due mainly to a further dilu-
tion of the heavy water by cellular body water. From these find-
ings it follows that all water molecules present in the plasma pass
with a very high speed through the capillary walls and with a
slower, but still remarkable speed through the cell walls, and \dce
versa. To this conclusion one may possibly object that heavy
water (DjO) may show a different behaviour from H-0 and recall
the results obtained in the investigation of the rate of hemolysis
of erytrocytes of cattle and rats, which was found to take place
about 44 per cent more slowly in DjO than in H-0 (Parpakt,
1936; Bkooks, 1935).

Contrary to those of the above mentioned authors, our experi-
ments were not carried out with pure DjO but with very diluted
heavy water, the viscosity and other properties of which only
slightly differ from those of H-O. Our most concentrated samples
contained, in fact, less than 2 per cent D-O, most of them contain-
ing very much less. The D-O injected into the vein is diluted at
once. The above mentioned authors did not use heavy water as
an indicator for water; they were interested in the differences
shown by H 2 O and DjO when penetrating into corpuscles. When
heavy water is used as an indicator, it should always be used in
a state as diluted as possible, partly for the above reasons and
partly because such diluted solutions contain mainly DHO
which is very similar to HjO, while D 20 is much less so.

PASSAGE OF WATER THROUGH CAPILLARY AND CELL WALLS. 17

In view of the high speed of capillary passage found in our
experiments, it is of interest to calculate the time taken by the
diffusion of water molecules through the capillary spaces. The
mean displacement r of water in water is \2 D (where D denotes
the diffusion constant of water in water, determined by using
heavy water as an indicator), is 2 cm per day (Orr and Thom-
son, 1935). Taking the size of the capillary vessel as 20 [i, we arrive
at the result that the displacement of water molecules within that
space takes 1 X 10-® day, or about 1/10 sec., thus an exceedingly
short time.

Summary.

Heavy water is injected into the vein of rabbits and blood
samples taken at intervals from the artery. The density of the
water prepared from the blood samples is determined and, from
the density difference between the injected heavy water and the
blood water, the extent of dilution, which the heavy water molec-
ules experienced in the body at different times, calculated.

As soon as a Ys min, after the injection, a dilution of the heavy
water by an amount of body water corresponding in volume to
about that of tbe extracellular space of the body, is foimd. This
very rapid rate of dilution is followed by a somewhat slower dilu-
tion process, in which presumably the cellular water participates.

After the lapse of less than Ys hour, tbe heavy water molecules
are evenly distributed over almost the total body water.

After the lapse of 39 days, only about Ys of the heavy water
injected is still present in the body.

There is no reason to assume that tbe heavy water (mainly
DHO) molecules show a markedly different behaviour from that
of the normal water (H.O) molecules present in the body, and we
have, therefore, to conclude that within about Ya min. a sufficient
flow of water takes place through the capillary walls to lead to an
almost perfect mixing of the blood and the interspace water.
An analogous interaction between cellular and extracellular water
takes less than Ya hour.

Beferences.

Brooks, S. C., J. Cell. Comp. Physiol. 1935, 7, 163.

Hj3VE.SY, G. and E. Hofer, jMature. Lond. 1931, 183, 495.
Linderstrom-Lang, K., 0. Jacobsen and G. Johansen, C. R. Lab.
CarLsberg. 1938, 23, 17.

2 — i01323. Acta phys. Scandinav. ToLI.

G. HKVIST AKD C. F. JACOSSEy.

Kr.CGH. A.. Acta Med. Scand., 1957, Sanplcment XC, 9.
ilcDoroALL, E. J.. F. Vkrs.^?., II. ErxEs^tKYER ar.d H. Gaertner,

Nature, L-ond. 1951, 1Z4, 1005.

O-ur., W, J. C. and J. A. V. Eutlep., J. Cbcm. See. 1935; 2, 1275.
Parpart, K., J. cell, cotnp. Phviiol. 1935, 7, 155.

U.-^sixa. H. H., Skandin. Arch. Physiol. 1955, 75, 225.

Aus dem biochemischen Institut der Universitat und der physiol.
Afat. des Karolin. Inst., Stockholm.

Znr Pliarmakologie der Cozymase."

Von

H. V. EULER, U. S. v. EULER nna F. SCHLENK.
Mit 5 Fignrcn im Text.

Die Cozymase, welche 1936 in fester Form dargestellt wurde
(H. V. Euler, Albers u, Schlenk), hat die Zusammensetzung
C 2 iHj,Oi«N 7 P. und wurde als Adenin-Nicotinsauxeamid-dinucleotid
erkannt (Schlenk u. H. v. Euler, 1936). Ihre Konstitution ist
nunmehr durch die Versuche von Schlenk (1936) auch in ihren
Einzelheiten aufgeklart*, sie wird dutch folgende Formel beschrie-
ben:

CONH,

j 0 :

HO OH I
(I;- C-C-C

H H H H

-0

C-

0 - P - 0

N = C - NHo

HC i-N

11 II

^CHr

1 _

1 / 1

i ^

N.-(

i - N - C

-C-

C-

6-c -

H H

Diese Formel stutzt sich auf den Nachweis, dass die Cozymase
einbasisch reagiert sowie auf die Isolierung und Charakterisierung
von Spaltprodukten, namlich

^ Der Redaktion am 3. Mai 1940 zugegangen. •, -nt*

* Die (ausserordentlich ■wahrscheinliche) Annahme, dasa die Pentose <ms Nioo-
tinsS-ureamid-nucleosidcs Ilibose ist, soli denmfichst noch besonders bestStigt
werden.

II. V. KIJLER, U. 6. V, EULER UND F. SCHLENK.

1) die Cliarakterisierung dcr Pentose-Pliosphorsaure, welche
von SciiLENK (1936) isolierfc nnd von Karrer (H. v. Euler,
ICarrer u. Becker, 1936) als Pentxjse-D-phosphorsaurc erkannt
wurde.

2) auf die Isolierung eincs von Vestin und H. v. Euler (1936)
dargestellten leicht hydrolysierharen Korpcr, nach Vestin u.
Mitarb. (1937) Adenosin-D-pjTophosphorsaure und

3) auf die Isolierung des Nicotinsiiureamid-nucleosides
(SCHLENK, 1938).

Cozymase kommt in fast alien Zellcn als unentbebrlicher Be-
standteil vor. Sic ist notwcndig fiir den Kohlenbydratabbau als
Wasserstoffubertriiger (H. v. Euler, Adler u. Hellstrom, 1935)
und zwar als Codeliydrase bei enzymatischen Teilrcaktionen der
Atmung, Glykolyse und Giirung, in erster Linie fiir die dicsen Vor-
gangen gemeinsame Reaktion:

Triosepbospborsiiure -|- Go Pbospboglycerinsaure + CoHj,
■vvobei die Cozymasc, Co, in Bihydrocozymasc, CoHj, iibergefiihrt
■wild.

Wie die Modellversucbe von Karrer (Karrer u. Mitarb.,
1937) gezeigt babcn, ist die wasserstoffiibertragcnde Fahigkcit der
Cozymase an die Pyridinium-Eorm gebunden, also an das Vor-
kommen einer Gruppierung (a), •vvelchc in (b) iibergeht:

(a)

= CH OH

-ca"

+ 2H

= CH

N-R (b)
- CH^

Die zweite Nucleotid-Komponente, die Muskeladenylsaure
(= Adenosin-5-pbospborsaure), geht, wie die obige Formel zeigt,
in das Molekiil der Cozymase ein.

Zur Phnrmakologie der Cozymase.

tiber die pharmakodynamiscben Wirkungen der Cozymase liegen
bereits Untersucbungen vor, die aber mit teilweise unreinen Pra-
paraten ausgefiihrt worden sind. Gard (1931) fand, dass ein
Cozymasepraparat von etwa ACo 100,000 den Blutdruck des
Kaninchens senkte und die Koronargefassdurcbstromung am
iiberlebenden Herzen erbobte. Die Wirkungen waren aucb nacb
Inaktivierung der Cozymase vorbanden. U. S. v. Euler und

ZUR PHARaiAKOLOGIE DER COZYMASE.

Gaddum (1931) unterwarfen ein Cozymasepraparat von etwa
demselben Reinheitsgrad einer etwas eingelienderen Priifung und
verglichen dabei die Wirkung von Cozymase niit der von reinem
Adenosin mit IPinblick auf die scbon damals vermutete Beziehung
der Cozjnnase zur Adenylsaure. Sie fanden, dass die Cozymase
etwa dieselben qualitativen Wirkungen auf Blutdruck und isolier-
ten Darm des Kaninchens soude auf isolierten Meerscbweincben-
utenis wie Adenosin zeigte. Kurz vorber batten Drury und
Szent-Gyorgyi (1929) auf die pharmakodynamiscben Wirkungen
der Adenylsaure und des Adenosins aufmerksam gemacbt. In
quantitativer Hiusicbt verbiclten sicb die Wirkungen von Cozy-
mase und Adenosin wie etwa 2 : 3, was mit den verscbiedenen
i\IolekuIargewicbten in Beziebung gebracbt wurde.

Mit einem reinen Cozymasepraparat baben wir die Wirkungen
am Herz, Blutdruck und Atmung des Kanincbens, sowie den Ef-
fekt am isolierten Kanincbendarm und Meerscbweincbenuterus
studiert.

a) Wirkung auf Herzfrequenz, Blutdruck und Atmung
des Kaninchens.

In i\Iengen von 0.2 — 2 mg der reinen Substanz ergab sicb nacb
intravenoser Darreicbung eine reine Drucksenkung wie in den
Abbildungen 1 und 2 zu erseben ist. Im Vergleicb mit Adenosin
(Brit. Drug Houses) war die Wirkung der Cozymase etwas ab-
weichend, was sicb einerseits in einer langsamer einsetzenden
Senkimg und in einem kleineren Betrag derselben darstellte. Mit
gleicb grossen Mengen von Cozymase und Adenosin (1 mg) be-
trug die maximale Drucksenkung des ersten etwa 30 % in 15
Sekunden, wabrend die entsprecbende Wirkung von Adenosin
40 % in 10 Sekunden war. Die Herzfrequenz nabm mit beiden
Substanzen betracbtlicb ab, von 340 bis 280 pro blinute nacb 1 mg
Cozymase und von 340 bis 250 nacb 1 mg Adenosin in einem Falle.
Das Blinimum der Herzfrequenz lag zwiscben etwa 5 und 15 Se-
kunden nacb der Injektion und die angefiibrten Zablen bezieben
sicb auf diese Periode.

Die Atmungsbewegungen wurden in den Blutdrucksversucben
registriert und zeigten regelmassig eine Zunabme von Amplitude
und Erequenz, wabrend der Drucksenkung. Es erscbeint jedocb
zweifelbaft ob diese Wirkung spezifiscb ist; sie kann auch eine
Folge der Drucksenkung sein.

II. V. KULER, U. 8, V. EDLKR END P. 8CHLEKK.

Abb. 1. Bhitdruck. Knninrben. Utr-th.in. Obcro Kiirve Alrmjnfrpltcwcgungcn.
Bci n 1 mg Adcnoxin, bej b I mg Coiymai'e intravcnOp.

Die Blutdrucksenkung nnch Adeiiosin und Adenylverhindungen
ist tcils auf cine Hemvirkung und teil.'i auf cine Gefasswirkung
zuriickzufiihren; bei dcr letztercn diirfte sowobl eine Konstriktion
der Lungengefasse Tsde cine Dilatation dor Artcriolen in mehreren
Gefassgebietcn mitw-irken (Drury u. Szekt-Gyorgyi, 1929;
Beknet ti. Drury, 1931; Drury, 1932; Zipp u. Giese, 1933).
Die erstgenannten Autoren geben aucb an, dass die Wirkung mit
der Desaminierbarkeit in Beziebung stebt, denn Hefeadenylsaure
wde Guanylsaure -n-aren inaktiv, und diese Verbindiingen -n'crden
nacb Schmidt (1928) nicbt oder nur langsara desarainiert. In
spateren Untersuchungen bat es sicb berausgestellt, dass Hefeade-
nylsaure eine Blutdruckwirkung besitzt (Bennet u. Drury;
Euler, 1933); diese ist jedocb von derjenigcn dcr Muskeladenyl-
siiure oder des Adenosins recbt verscbieden und verlauft viel lang-
samer. Die aucb von Embden (1932) gebegte Auffassung, dass die

ZDR PHARMACOLOQIE DER COZYMASE.

Abb. 2. Blutdruck, Kaninchen, Urethan. Die verschiedenen Blutdruokkurven
zeigen die Wirkungen von 0.2, 0.5, 1 und 2 mg Cozymase intravenos.

Desaminierbarkeit von grosser Bedeutung fiir die Aktivitat ist,
vrird von Parnas und Ostern (1932) bestritten.

Wir baben die Blutdruckwirkung von einigen Adeninderivaten
im Kaninchenversucb gepruft (Abb. 3). Hieraus ergibt sicb dass
Inosinsaure und Desaminocozymase vollig un-wirksam waren,
vabrend Hefeadenylsaure, Adenosin und Adenosintripbospbor-
saure, wie scbon friiber bekannt war, blutdrucksenkend wirkten.
Bedeutende quantitative Verscbiedenbeiten in der Wirkung waren
zu verzeicbnen; somit war die Wirkung von Adenosin, auf gleicb
grosser Menge bezogen, viel starker als die der Hefeadenylsaxure,
und Adenosintriphospborsaure nocb bedeutend wicksamer als das
Adenosin. Besonders auffallig ist der Unterscbied in der Wirkung
zwisc ben Hefeadenylsaure und Adenosin, die viel ausgesprocbener
ist a Is in den entsprecbenden Versucben von Bennet und Drury.
Es liegt die Vermutung nabe, dass die Verscbiedenbeit der Wirkung
nait der Zusammensetzung der Praparate in Beziebung stebt.
Das in unseren Versucben verwendete Adenosin stammte von
British Drug Houses und die Hefeadenylsaure von Boebringer
u. Sobne. Die Desaminocozymase (Schlene, Helestrom u.

II. V. EULI:R, U. B. V. EDLUR UN» P. EOfILENK.

Abb. 3 . BluWnick, Kaninchcn, TJrethnn. A. 2 mg Bc-wminocozyinase, B. 1 mg
InoBinsIiurc, C. O..'* mg HcfeadcnylsSurc, D. 0.5 mg Adcno^m, E. 0.5 mg Adcnosin-
tripliospborsitirc, intravcnOs.

H. V. Euler, 19.38) in v.'e!cher die NH.-Gruppe dcs Adenin-
restes, •ft'ie in dcr Inosinsilnrc, dutch OH ersetzt ist, wurde
von F. ScHLENK dargcstcllt.

Unsere Vcrsuchc haben cine weitere Stiitze fiir die von Drury
u. Szel'T-Gyorgyi und Embden nu.sgesproc}iene Mcinung beziiglich
der Bcdcutung dcr Dcsaininierbnrkcit gelicfert, vrenn nuch andcrc
Erkliirungsgrundc fiir die Vcrscliiedenlicifc der 'Wirkung in Bc-
tracht kommen konnen.

B) Wirkung an isoliorton Organon.

Am isolierten Kaninchendnrm %virkt die rcine Cozymase hem-
mend in cincr iihnlichen Weisc wie das Adenosin. Ein quantita-
tiver Unterschied zvischen den beiden Substnnzen bestand auch
hier, indem die Wirkung von Cozymasc in einer Ivonzentration
von 1 : 150,000 sclnvacher war als die von Adenosin in gleicher
Konzentration. (Abb. 4 A.)

Wiihrend keine der beiden Substanzen in einer Konzentration
von 1 : 30,000 auf den isolierten Kaninchenuterus wirksam var,

konnte mit beiden cine Kontraktion des Sleerschweinchenuterus
erzielt werden. In diesem Falle waren beide Substanzen etwa gleich
stark wirksam, vrie aus der Abb. 4 B hervorgeht. Each Vcrsuchen
von Deuticke (1932) steigt die Uteruswirksamkeit mit steigender
Phosphorylierung des Adenosins.

Ausser der Cozymase haben wir ebenfalls Desaminocozymase,
Inosinsaure, Hefeadenylsaure und Adenosintriphosphorsaure am

ZUR PHAKMAKOLOGIE DER COZYMASB.

Abb. 4. A. Isoliertcr Kaninchcndnrm. n, c, und o Adcnosin 1 : 150,000, b. Co-
zymnsQ 1 : 160,000, d Cozymase 1 : 00,000. B. Isoliertcr Jlecrschweinchenuterus.
a CozjTJinse 1 : 00,000, b. Adcnosin 1 : 150,000, c Adenosin 1 : 00,000.

Kaninchendarm gepriift. Abb. 5 zeigt, dass auch Her die Wirkung
der Hefeaden 3 ’’]saure treit hinter der des Adenosins steht. Dagegen
war die "Wirkung der Adenosintriphospborsiiure — im Gegensatz
zur Blutdruckwirkung — schwaclier als die des Adenosins. Inosin-
saure, •wie auch Desaminocozymase waren in den verwendeten
Mengen unwdrksam.

Zusaminenfassend lilsst sich sagen, dass die reine Cozymase in
grossen Ziigen die charakteristischen Wirkungen einer Adenylver-

Abb. 5. Isolicrtcs DiinndarmstOck, Kaninchen, Badvolumcn 30 ccm. A. 0.6 rog
Hcfcadenylsaurc, B. 0.6 mg Aticnosin, C. 0.5 mg Adcnosintriphospbors5urc, D.

0.5 mg Desaminocozyraasc, E. 0.5 mg Ino^iosaare.

bindung zeigt, wean auch gewisse Verschiedcaheiten sowohl in
qualitativer -wie quantitativer Bichtung beobacbtet warden
konnten.

Besprechung.

Adenylsaure und Adenosin-triphospborsaure (wie auch Adeno-
sin-diphosphorsaure) spielen eine ausserordentlich ^dchtige Rolle
als Phosphatiibertrager (Lohmank, 1931). Co zymase selbst ist
als Phosphatiibertrager in Betracht gezogen worden (auf die
diesbzgl. Diskussion soil hier nicht eingegangen werdeh) (Loh-
MANN u. Meyerhof, 1934). Cozymase steht jedenfallsder Adenyl-
saure in dieser Hinsicht an Wirkungsfahigkeit nach. Man hat
gelegentlich angenommen, dass ein bewegliches enzymatisches
Gleichgewicht statthat:

Cozymase (Dihydrocozymase) Bicotinsaureamid-nucleotid
4- Adenylsaure. Dafitr, dass ein solches bewegliches Gleichge-
wicht eine physiologische Eolle, im Organismus spielt, fehlen aber
bis jetzt ausreichende Anhaltspunkte; man wird fur die Koppelung

ZUR PHARMAKOLOQIE DER COZYMASE.

zwischen Wasserstoffiibertragung und Umpliospliorylierung bei
der Glykolyse cine andere Erklarung suchen miissen.

Fiir die Beurteilung der pbarmakologischen Wirkungen der Co-
zjinase, der Adenylsaiire, Adenosintriphospliorsaure und des
Adenosins sind folgende Tafsaclien in erster Linie in Betracbt zu
ziohen.

a) Adenosin und Cozymase zcigon angeniiliert die gleicbe blut-
drucksenkende "Wirkning am Kaninchen und hemmende Wirkung
am isolierten Kanincbendarm. Die Blutdruckn'irkung der Ade-
nosin-tripbospborsaure ubertrifft dicjenigc der Cozjunase und der
anderen Adcnosinderivato.

b) Adenin ist wirkungslos, ebenso Desaminocozymase, Inosin-
saure, Guanylsiiure und Nukleinsiiure.

c) Hcfeadenylsaure {Adenosin-3-pbospborsaure) ermes sicb
bei unseren Versuchon (im Gegensatz zu denen von Paestas und
OsTERX, 1933) als venigcr nirksam. Die gefundene sclnvacbe
Wirksamkeit durfte darauf zuriickzufuhren sein, dass die Hefe-
adenylsaure beini Yersucb depbosphoryiiert wird, worauf Adeno-
sin zur Wirkung kommt.

Die Phospborylierung des Adenosins verlauft, soviel bekannt ist,
im Tierkorper rascb; das gleicbe gilt von der Spaltung der Cozy-
mase, (H. V. Euler u, Heiwixkel, 1937; H. v, Euler, Sohlenk,
HEnvixKEL u, Hogberg 1938). Das System Adenylsaure-
Adenosin-dipbospborsaure und Adenosin-tripbospborsauxe liefert
das Coenzym der Phospborylierung und ist dadurcb mit dem
anaeroben und aeroben Koblenbydratabbau verkniipft.

Die bier gefundenen imd erwabnten Tatsacben berecbtigen zur
Annabme, dass Cozymase vrie auch Adenosin und dessen blut-
drucksenkenden Derivate in dem Masse vrirken, als sie durcb
Spaltung bzw. Synthese in Copbosphorylasen iibergeben und so
den Phospbatstoffu'echsel bescbleunigen.

Zusaramenfassnng.

1. Heine Cozymase (Adenin-nikotinsaureamid-dinukleotid) be-
sitzt annahernd die gleicbe blutdruckssenkende, darmbemmende
und uteruserregende Wirkung Trie Adenosin. Desaminocozymase
ist in entsprecbenden Mengen unwirksam.

2, Hcfeadenylsaure vrar in unseren Yersucben vreniger Tvirk-
sam als Adenosin, das seinerseits scbTracber als Adenosin-tri-
pbospborsaure ndrkte.

n. V. KL’I.KH, U. K. V. KULKU l.’.VD I’. FCrtLKKir,

3, Din Wirktinn tk-r CJozynuiuo utirl flnr Aflnnylvnrbindungnn
win! luit ilirnr Dn/}n!j(!in« fiir /Jnn PJiofipkalfTt^if^rnchKe! in Ik-
zinliuii}' p‘.Hn(?.t,

Lttorntur.

D. W. mu! A. N. I)w;r.v, .T, I’hyi-iol. IMl. 72, 2“-?.
DKrvicKi;, 11. .1., rfliif*. Arnb. p"', riiv;*!nl, lb.32, 220, 'i37.

DtutUY, A. X., .T. I’liNvitil., 7f, i\l.

— uih 3 a. .S?j;NT'nyonnvJ, Kinmin., Ik 22 , CS\ 233 .

I'kniPK.v, 0., Z. nn.';. Climn., 11*32, 40, 331.

ICri-Kr., TI. V., Ik Aitt.ni'. tititl II. liru^Tnrut, Sy. Vet. Tifkkr,, 11*33,
■it, 2:n>.

— , II. At.iu:jt« untl F. Srnu:,*;f:, Jlopj'n-S'yl, Z., 32.33, 240, 113,

— nntl 13. Hf:nvi.s'Kr;i,, 11*37, 17, 231*.

— , P. uiifl PycKJ’P., IIolv. Cfjim. Acta, I2.3*i, JO, lOvO,

F. Hciiu:nk, H. IfKivvi.\-Ki:i. un'l 15. Ik^tiuinr;, Kl.nn'1a 31*3''',
20G, 2‘3't.

Kui.r.n, U. S. v., Ardt. nrp. Pntb. Pbanii.ik. 1232, 107, 171.

— unri J. 11. (JAltDi:?!, .3. PliVMi>l, 12.31, 72, 7».

o.vnu, s., no[>p>-Snv!. z. vxn'ioo, cr,.

II.Ar.TM.vN.N-, Z. klin.'Mnd., 12.32. 121, .121.

KAr.r.r.n, P. mid Mitarb., Hdv. Chirn. .-Vetn 11*30, 19, .Sil.

— , K bond a 3 2.37, 20, 59.

Lon.M.A.v.v, K. Piiodicm. Z. ll*,31,p.//, OT.

Kbnndn 11*31, 271, 20 1.

~ und O, .MnvtrnnoK. Kbnrid.n 1931, 272, f>0.

P,\r.N.\s, J, K. mid P. O.’TKi'.v, Klin. W?dir. 19.32, 11, 1.331.
P.M’.-‘-CTfi.v, C. ]{. Ao.nd. f?ci. 19.31, .3, 020.

.S(mu:sK, F., -Uk, f. Komi. 19.30, 12 11, Xr. 17.

— , Xatnnvb.'innrcbaltnn 1910, 22, 40.

— und II. V. Euum, Kbnnd.a 19.30, 21, 791.

— , IIeu.hthOm, II. mid H. v. Bor. d. d. diem. Go?. 193S,

71, 1171.

GuNTHEit, und II. V. Eui.t;p.. Ebnnda 193S, 12 7 :, 30.

FcnMTDT, a., nopi)c-.Scyl. Z. 192d, 179, 243.

Vestin', B. und il. v, EutKii, Ebmd.'i 193*3, 215, 1.

— , .‘rcntitNK, F. und II. v. Eveer, Ber. d. d. diem. Ges. 1937,
70 , 1.3G9.

Ztpf, K. und W. Gir.sr., Arch. exp. Path. Phnrin.ak. 19.33, 171, 111.

From the Institute of Medical Chemistry, University of Upsala,

Su'eden.

Studies ill Glycoproteins.^

QUNNAR BLIX.

("With 2 fignrcs in Iho text.)

In the past glycoproteins have usually been defined as con-
jugated proteins in -which a simple protein is combined with a
carbohydrate compound other than nucleic acid. As in later
years simple proteins such as albumins and globulins have been
found also to contain a carbohydrate group, this definition has
become unsatisfactory. Between the proteins hitherto called
simple proteins and those comprised xinder the name glyco-
proteins there exists, however,- a certain quantitative difference
vdth regard to the carbohydrate group. The former contain at
most a few per cent of carbohydrate, whereas the carbohydrate
content of the latter amounts to 10 — 20 per cent or even more.
This fact justifies the keeping of the name glycoproteins for the
substances until now grouped under tliis name. It must be re-
membered, ho-wever, that no sharp differentation between simple
proteins and glycoproteins can be based on the difference just
pointed out, since some proteins have recently been found which
contain 6 — 10 per cent of carbohydrate.

There is a small group of glycoproteins -which contain a neutral
carbohydrate component of the same type as that of the simple
proteins, i. e. a polymerised dihexose-hexosamine. The ovo-
mucoid and the seromucoid are the best kno-wn members of this
group. It is worthy of notice that these substances differ from
all other glycoproteins in that they are not precipitated from their
neutral solutions on the addition of acetic acid. — It is generally
believed that the carbohydrate in these mucoids as well as that
in simple proteins is firmly bound to the protein. This opinion

^ Eeceived for publication 7 May 1940.

GUNEAB BLIX.

appears to rest almost exclusively on the fact that an isolation
of the carbohydrate component has not been possible to attain
without a far-going disintegration of the protein. Although not
quite conclusive this fact no doubt gives strong support to the
opinion mentioned. Direct evidence of the ovomucoid being a
chemical imit and not a mixture of carbohydrate and protein
was recently given by Hesselvik (1938) in this laboratory, who
found this protein to he electrophoretically well-defined and
quite uniform.

With the exception of the mucoids of the group just mentioned
all other glycoproteins seem to contain an acid carbohydrate.
This is true for the mucins, such as submaxillary and synovial
mucin, as well as for tissue mucoids, e. g. chondromucoid, tendo-
mucoid and . hyalomucoid. The nature of the carbohydrate
group in these substances is not in all instances clear. For a long
time the generally accepted view of the carbohydrate component
was that of Levene (1925): that it was constituted either of chon-
droitin sulphuric acid or mucoitin sulphuric acid. The chondro-
mucoid no doubt contains chondroitin sulphuric acid. The same
carbohydrate acid seems also to be present in all mucoids of the
connective tissues, including those of bone and umbilical cord
(Meyeb 1938). The mucoitin sulphmic acid, on the other hand,
has actually never been demonstrated as being the main carbo-
hydrate component in any isolated glycoprotein. Levene and
his co-workers (1925) prepared mucoitin sulphuric acid from
various materials containing glycoproteins such as gastric mucosa
and cornea but not from isolated glycoproteins. Glycoproteins
in these materials are rich in carbohydrate but contain so little
sulphuric acid, that mucoitin sulphuric acid, if actually present,
probably ought to be regarded more as an impurity than an
integral part of the protein (Blix et al. (1935), Blix (1936),
Karlbeeg (1936)). Ester bound sulphuric acid has, in, fact,
never been found in more than negligeable amounts in other
glycoproteios than those containing chondroitin sulphuric acid,
i. e. chondromucoid, tendomucoid etc.

Two new acid carbohydrates, both sulphur-free, have in later
years been demonstrated as components of glycoproteins. One
of them was first prepared from vitreous body and therefore
named hyaluronic acid. According to Meyer et al. (1934, 1936) it
is composed of equimolar parts of glucuronic acid, glucosamine and
acetic acid, and might therefore have been formed from mucoitin

STCDIKS IN GLYCOPROTEINS.

sulphuric acid through splitting off the sulphuric acid. As hyalu-
ronic acid can be obtained in the pure state from different materi-
als employing only very mild procedures, it is, however, no doubt
pre-formed in the living tissues. Besides in the vitreous body
the hyaluronic acid has been found in synovial mucin, in umbilical
cord and in aqueous humor (Meyer (1938)). — In submaxillary
mucin Blix (1936) found a carbohydrate acid, composed of hexos-
amine, acetic acid and a hydroxacid. The nature of the latter
has not yet been elucidated. Possibly it might be a desoxyhex-
mronic acid (Blix (1938)). This carbohydrate thus seems to be
chemically closely related to the hyaluronic acid.

As the protein component in the glycoproteins containing an
acid carbohydrate can be supposed to contain, just as other
simple proteins, some neutral carbohydrate, these glycoproteins
should indeed contain two different carbohydrates, namely, in
addition to the main, i. e. the acid one, a few per cent of a neutral
carbohydrate. This has actually been demonstrated for the sub-
maxillary (Blix (1936)) and for the synovial mucin (Meyer
et al. (1939)).

Concerning the relations between the carbohydrate acids and
the proteins accompanying them in Imng materials the knowledge
is still very incomplete. Morner (1889) and Sohmiedeberg
(1891) were of the opinion that the chondroitin sulphuric acid
was partly present as alkali salt in the cartilage. Morner (1889)
showed that chondroitin sulphuric acid forms precipitates with
simple proteins when neutral solutions of these substances are
acidified with acetic acid. Such precipitates are generally formed
between chondroitin sulphuric acid and proteins on the acid
side of the isoelectric point of the latter, that is, between the
carbohydrate acid and protein cations. The same appear.'- to
be true for the hyaluronic acid (Meyer et al. (1936)). As the rou-
tine method for the preparation of most glycoproteins containing
carbohydrate acid consists of extraction under neutral or faintly
alkaline conditions followed by precipitation by addition of acetic
acid, it is tempting to think that the glycoproteins thus precipi-
tated are artificial products, the components existing uncombined
with each other in the tissues as well as in the neutral or alkaline
extracts, combining only when protein cations are created on
acidification. Only with a basic protein such as histone should
the carbohydrate acid be able to combine in this way at hydrogen
ion concentrations within the physiological range.

QDNNAK BLIX.

Tlie supposition that the chondroproteids should be regarded
as salts between carbohydrate acid and protein was put forward
many years ago (Sohmiedeberg (1891), Takahata {1924)).

Meyer et al. (1937, 1) have found the complexes formed be-
tween chondroitin sulphuric acid and simple proteins remarkably
constant in spite of wide variation of the relative amounts of the
reactants used. The acidbinding capacity of the protein compo-
nent was found to be about the same when calcxdated from the
carbohydrate-protein complex (taking the carbohydrate as a
dibasic acid) as when calculated from compounds between pro-
tein and acid dyes or from the basic amino acid content of the
protein. The ratio between hexosamine and nitrogen in native
cartilage was found to be about the same as in artificial chondroitin
sulphuric acid-gelatin complexes. SIeyer et al. (1937, 2) there-
fore consider “the major part” of the cartilage to be a protein
salt of chondroitin sulphuric acid.

Recently Meyer (1938) suggested that also the glycoproteins
derived from vitreous humor, umbilical cord and synovial fluid
are to be regarded as salts between polysaccharide acid and pro-
teins and that in the native materials the hyaluronic acid exists
"largely in dissociated form”. This view appears to be based
chiefly on the facts that the hyaluronic acid can be isolated from
the mentioned materials -without the use of hydrolytic means and
that the artificial glycoproteins prepared from hyaluronic acid
and simple proteins are very similar in composition and properties
to the glycoproteins obtained from various natural sources. —
Meyer even proposes that the term “mucin” should be used
“only in a physiological sense to denote a viscous fluid of secre-
tory origin”.

It is clear that if compounds between proteins and carbohydrate
acids exist in Imng tissues, the binding between the components
must be a loose one, at least in the instances where they have been
separated -without hydrolytic procedures. The polysaccharide acids
might exist in salt-like combination with basic proteins inira
vitam, but as yet practically nothing is known about the electro-
chemical properties of the protein in the materials in question.
It should not be overlooked that loose complexes of another kind
between protein and carbohydrate acid may also occur. Direct
experimental e-vddence of the relations between protein and
carbohydrate acid in native materials is obviously still very
limited.

STUDIES IN GLTCOPKOTEINS.

The aim of the present investigations has been to find out by
means of electrophoretic analyses if, in solutions of some typical
glycoproteins containing a 'polysaccharide acid, or in native ma-
terial holding such glycoproteins, the carbohydrate exists free
or in combination with protein. The glycoproteins chosen for this
investigation vere the folIoAving; hyalomucoid, syno^dal mucin
and submaxillary mucin. In addition a few experiments were
conducted on cartilage extracts.

Before entering upon a description of the experiments performed
it is necessary to give a short account of some electrophoretic
investigations recently made on vitreous body and on synovial
fluid.

Hes-SELVIK (1938) submitted filtered vitreous bodies to electro-
phoresis in the apparatus of Tiselius. Three distinct boundaries
regularly appeared, the mobilities of which were determined at
various pH values. The fastest migrating boundary (u about
12 cm- • volt“i -sec."^ X 10^ at pH 7.5) disappeared if the mucoid
was removed before the electrophoresis by precipitation with
acetic acid. The electrophoretic mobilities of the other two
components were of the same order as those of albumin and
y-globulin of blood serum. It was concluded that the fastest
migrating component was the hyalomucoid. As, however, this
component was not isolated and analysed, it might have been
free hyaluronic acid. The synovial fluid has been electrophoretic-
ally studied by Hesselvik (1940). It shows four distinct bound-
aries. The fastest boundary (u about 11 cm= • volt“^ ‘sec."^ X 10®
at pH 7.5) disappears if the mucin is removed by precipitation
with the aid of acetic acid. The mobility values of the other
three correspond to those of serum albumin and serum globulin p
and y respectively.

The submaxillary mucin has, as far as I know, not been previ-
ously investigated electrophoretically, nor have cartilage extracts.

Experiinentnl.

The electrophoresis method recently described by Tiselius
(1937) was employed. The glycoprotein solutions were dialysed
before the electrophoresis against a phosphate buffer of pH 8.03
and of a ionic strength of 0.1. The same buffer was used as
electrode solution. The electrophoreses were carried out at J:; 0°
and with a potential gradient of about5V/cm. — The nitrogen

3 — 'i01323. Acta phys. Scandinav. Yol.I.

GUNNAE BLIX.

determinations were made •witli the micro Kjeldahl method.
Hexosamine ■was determinated by the method of Elson and
Moegan as modified by Nilsson (1936). 2 ml. of the solutions
analysed "were heated for 12 hrs. with 5 ml. 2 N HCl on the boiling
water bath before the hexosamine determination.

1. Experiments with -vdtreous body and with hyalomucoid. —

500 ml. filtered •vitreous bodies from cattle were dialysed against
water, concentrated at room temperature to 100 ml., filtered
again and dialysed overnight against the buffer used in the
electrophoresis experiments. In agreement with earlier findings,
three boimdaries appeared on electrophoresis. The fastest
migrating component was isolated in a series of experiments.
The yields from all experiments were collected and the solu-
tion (28 ml.) after concentration to half the volume, sub-
mitted to a new electrophoresis. The second electrophoresis was
performed in order to free the solution from traces of the other
components which might have been present on account of a not
quite perfect separation. On this occasion two boundaries ap-
peared on each side. The faster one (fraction I) showed the same
migration velocity as before, the other (fraction II) did not
move appreciably during the time of observation. The two
fractions were separa'ted and analysed. “Fraction II” was quanti-
tatively quite -unimportant. As the ratio between nitrogen and
hexosamine was the same in both fractions, the slower boundaries
are believed to have been anomalous (6- and e-effect).^ Such
anomalies are liable to occur in electrophoresis experiments "with
rapidly migrating substances. This opinion is supported by the
fact, that the “schlieren” bands of these two boundaries differed
markedly, that on the cathode side being much weaker and more
diffuse than that on the anode side.

Fraction I gave a negative biuret reaction. It contained:

Hexosamine 0.412 mg./ml.

Nitrogen {olwe?} mg-/ml.

Calculated from the hexosamine content the solution should
contain 0.914 mg. hyaluronic acid and 0.032 mg. nitrogen per

^ See for example Tiseuvs and Kabat (1939) and LoifOSWOETH and Sic
iNiiES (1939).

STUDIES IN GLYCOPROTEINS.

ml. A small nitrogen excess is thus found. Calculated as protein
it should, however, amount to only about 10 per cent of the
hyaluronic acid present. Ob\dously then the fastest migrating
component of the vitreous body is not glycoprotein but pure
or almost pure hyaluronic acid.

In another experiment the hyalomucoid was precipitated in the
ordinary way from filtered ^^treous bodies, redissolved with the
aid of alkali and the solution obtained electrophorized. Two
boundaries appeared, the faster (F I) showing the same migration
velocity as the fastest component in the vitreous body fluid, the
slower (F II) migrating with about the velocity of serum albumin.
Both components were separated and analysed. F I contained:

Hexosamine 0.42 mg./ml.

Nitrogen 0.076 mg./ml.

The nitrogen excess was slightly greater than in the foregoing
experiment. '\Micther or not this was due to contaminating
protein could not be decided. The biuret reaction of F I was
negative. F II contained:

Hexosamine 0.04 mg./ml.

Nitrogen 0.51 mg./ml.

The biuret reaction was positive. This component was thus a
protein containing about 1.2 % hexosamine.

h. a. alb. —

Fig. 1* Concentrated vitreous body fluid. — h. a. = hyaluronic acid.

Fig. 1 shows an electrophoretic diagram of concentrated vitreous
body fluid, taken according to the inclined slit method recently

GUNNAR BLIX.

developed by Sveksson (1939, 1940). It sbould be noticed that
the peak for the hyaluronic acid is completely separated from
the protein peak next to it, as this provides evidence against the
existence of a loose, dissociable compound between the carbo-
hydrate acid and the protein.

Evidently, then, the hyalomucoid dissolved under neutral
conditions is no chemical unit but consists of two apparently
uncombined components, viz. hyaluronic acid and a simple
protein. The hyalomucoid is an artificial compound, formed on
acidification of neutral solutions containing hyaluronic acid and
vitreous body protein. In the vitreous body the hyaluronic acid
is present, probably exclusively, in the form of salt with inorganic
bases and not combined with protein.

2. Experiments on synovial fluid and on synovial mucin. —

Synovial fluids from horses were dialysed overnight against
water and precipitated with acetic acid. The mucin obtained
was redissolved with the aid of alkali, care being taken that the
pH resiilting from this procedure did not exceed 8. The mucin
was precipitated and redissolved in this way four times. Its
characteristic physical properties were not changed during this
treatment. The solution finally obtained was dialysed against
the buffer used in the electrophoresis. Two boundaries appeared.
The faster had a mobility value of 12 cm- • volt*** ’ sec.~^ • 10' at
pH 8.03. The faster component of the native synovial fluid
shows about the same mobility. In a series of electrophoresis
experiments the faster component was separated. In all, 28 ml.
of solution containing this component were collected, and after
reduction to half the volume submitted to a new electrophoresis.
This time also two boundaries appeared, the one rapidly migrating,
the other stationary. The non-migrating boundaries were to all
appearances of anomalous character. The solution of the migra-
ting substance contained:

Hexosamine 0.620 mg,/ml.

fO.0672
io.0736

The biuret reaction was negative. The test with sulfosalicylic
acid was likewise negative. Calculated from the hexosamine value
the solution should contain 1.375 mg, hyaluronic acid and 0,048
mg. nitrogen per ml.

I 0.070 mg./ml.

STUDIES IX GLYCOPROTEINS.

Here also a slight nitrogen excess was found. Calculated as
protein it should amount to only 1/10 of the hyaluronic acid
present.

In another experiment the slower migrating component of the
.synovial mucin was isolated electrophoretically and analysed.
The analyses showed:

Hexosamine 0.030 mg./ml.

Nitrogen 0.3S mg./ml.

The biuret reaction was positive. This component was thus an
ordinary simple protein. — In a third experiment synovial fluid
was dialysed against water for 24 hrs. and then precipitated
with acetic acid and filtered. The filtrate was analysed for
hexosamine and nitrogen. The hexosamine/nitrogen ratio found
corresponded to that for a simple protein containing 0.9 %
liexosamine.

Tig. 2 shows an electrophoretic diagram of synovial fluid.
*Vlso here the peak for the hyaluronic acid is completely separated
from the protein peak next to it.

Fig. 2. Svnovial ■flnid. — b. a. = hyalnronlc acid.

It is thus clear that the synovial mucin too is an artificial
product, formed only on acidif}dng the synovial fluid with acetic
acid. In the synovial fluid the hyaluronic acid appears to be
present solely in the form of salts with inorganic bases, and not
in combination with proteins.

.3. Experiments on submaxillary mucin. —

The submaxillary mucin Avas prepared according toHAMJiAESTEN
(1881). It was dissolved Avith the aid of hydrochloric acid and re-
precipitated by dilution with Avater, this procedure being repeated
four times. The mucin finally precipitated Avas Avashed free from

GUNNAR IlLIX.

HCI and tlien dissolved in water with the aid of NaOH at a pH
not exceeding 8. The solution was thereafter dialysed against
the buffer used throughout and submitted to electrophoresis.
Submaxillary mucin prepared in this way showed regularly two
distinct boundaries, symmetrically on 1)oth sides. In Table I
the migration velocities of these boundaries at different pH
values are given.

Talile I.

Mobilities in cm*.

volt. ‘ sec.“* X 10^ '

Faster component

; Slower component i

6.12

— 7.1

; 1

1 ±0.0 !

6.80

- 8.3

1 ±0.0 :

7.50

- 9.7

{ - 0.2 ’

8.03

-10.7

! -o.c ;

The velocity of the faster component (F I) is somewhat greater
than that of serum albumin under the same conditions but not
so great as that of hyaluronic acid. The mobility of the slower
component (F II) is of the same order as that of the y-glo-
bulin.

In a number of electrophoreses 28 ml. of F I in all were sepa-
rated. The solution was concentrated to half the volume, dialj'sed
against the usual buffer and once again submitted to electro-
phoresis. Two components appeared, whieh were separated
and analysed. The slower component was present in such small
amounts that the analytical values became uncertain. They are
therefore not given here. The solution containing the main, i. c.
the faster, component contained:

Hexosamine |o 524 ] mg./ml.

Nitrogen {a542| mg./ml.

The biuret reaction was positive.

In another experiment the slower component (F II) of the sub-
maxillary mucin was isolated and analysed. It showed;

Hexosamine 0.032 mg./ml.

Nitrogen 0.099 mg./ml.

STUDIES nr GLTCOPROTEnrs.

Evidently tlien tlie submaxilJary mucin as prepared in this
vork is electrophoTctically not quite uniform. In approximately
neutral solutions the main component is a glycoprotein containing
about 27 % poh-saccharide acid (mol. weight 381). A minor part,
approximately 10 % of the whole, is constituted of a protein
with a hexosaininc content of about 5 %. As this protein shows
a mobility of about the same order as the y-globulin, it is un-
likely that it contains an acid polpaccharide.^ It probably ought
to be regarded as a contaminant.

In one experiment submaxillary glands were repeatedly extract-
ed with water. The extracts still coloured by haemoglobin were
rejected. The first practically colourless extract was directly
submitted to electrophoresis. Two components appeared, showing
identical mobility values with those of the submaxillary mucin.
The faster one was separated and gave the characteristic colour
reactions for submaxillary mucin. It is thus clear that not even the
aqueous extract of submaxillary gland from which the mucin is
prepared contains any unbound polysaccharide.

4. Experiments on cartilage extracts. —

100 g. of cartilage (nasal septums from cattle) were extracted
at room temperature for two days with distilled water. The
extract was centrifuged and filtered. The filtrate was reduced
to half the volume at room temperature and dialysed against
the usual buffer. The solution was then electrophorized in the
larger apparatus of Tiseuius (1938). Two components appeared.
Both were isolated and analysed. In a special experiment their
electrophoretic mobility was determinated at pH 8 and found
to be 17.0 and 7.9 cm= • voltr^ • sec.“^ X 10® respectively. To judge
from this result the faster component was in all probability not
a protein. The chemical analyses of its solution gave:

Hexosamine 0.208 mg./ml.

Nitrogen 0.077 mg./ml.

The nitrogen excess is somewhat greater than for the hyaluronic
acid in the experiments with vitreous body and synovial fluid.
Calculated as protein it should amount to about of the total
substance (the non-protein part taken as chondroitin sulphuric

* The possibility cannot be excluded that by the separation of the slower
component some contamination with the faster one may have occurred. This
may account for the relatively great carbohydrate content of the former.

GUNNAR BLIX.

acid). However, the solution of the faster component showed
a negative biuret test and a negative sulfosalicylic acid reaction.
It is therefore believed that this component was almost pure
chondroitin sulphuric acid.

The slower component was a protein containing only 0.02 %
hexosamine.

With water alone only a minor part of the chondroitin sulphuric
acid in the cartilage can be extracted. Clearly then, the above
results support the old view of Mop.xer (1889) that the chondro-
itin sulphuric acid in the cartilage is partly present as alkali salt.

Bi.scussioii.

Two typical glycoproteins, the .synovial mucin and the hyalo-
mucoid have proved to be purcl}" artificial products. In the
synovial fluids as well as in the vitreous body the hyaluronic
acid is, to all appearances, present exclusively in the form of salts
with the coexisting inorganic bases and not in combination with
the proteins. It is not likely that the synovial mucin and the
hyalomucoid are unique among the glycoproteins in being artificial
products. However, the outcome of the electrophoretic experi-
ments on the submaxillary mucin should remind us of the danger
in generalising from a few cases. The submaxillary mucin seems
to exist also in neutral solution as a definite compound between
protein and carbohydrate. It is very unlikely that the carbohydra-
te and the protein should have migrated with the same velocity
at the different pH values if they had not been in some way
combined vdth each other. It remains to find out the nature
of this combination.

The improved knowledge of the chemistry of the glycoproteins
gives, in my opinion, no reason for using the terms “mucin” and
“mucoid” in an)’ other sense than that in which they have been
used in earlier years. “Mucin” should still signify animal ghjeo-
proiehis which occur in or may be prepared from mucus, mucous
fluids or mucous tissues, “JIucoid” should signify glycoproteins
occurring in or derived from other animal materials. It must be
remembered nevertheless that certain glycoproteins are purely
artificial products and that the peculiar mucinous character of
various materials of animal origin may, in some instances, be due
not to a glycoprotein but to a polysaccharide acid alone.

It would, at present, appear perhaps somewhat premature

STUDIES IN GLYCOPROTEINS.

to attempt a chemical classification of the glycoproteins, never-
theless, it is helicved that the classification given helov' may, as a
preliminary one, be of some value.

Glycoproteins.

I. Glycoproteins containing a neutral polysaccharide (Neutro-
glycoproicins). — Ovo- and sero-mucoid belong to this group.

II. Glycoproteins containing an acid polysaccharide^ {Acidd-
gjycoprotcins). This group includes the following sub-groups:

1. Chondroproteins, containing chondroitin sulphuric acid.
This group includes the chondromucoid and probabl}’" all mucoids
of the connective tissues.

2. II yah proteins, containing hyaluronic acid. Only artificial
hyaloproteins are as yet known.

3. Sialoproteins, containing the acid polysaccharide present in
the submaxillary mucin.

4. The term nnico proteins should be reserved for eventual
compounds between mucoitin sulphuric acid and protein.

Summary.

A short re\new is given of the results of recent investigations
on the chemistry of the glycoproteins and especially of the
present state of the question concerning the relations between
carbohydrate and protein in these substances.

Hyalomucoid, synovial mucin and submaxillary mucin were
investigated Avith the aid of electrophoresis. In approximately
neutral solutions the hyalomucoid and the synovial mucin proved
each to contain two electrophoretically Avell-defined components.
In both cases the faster migrating component was found to consist
of a pure polysaccharide acid (hyalui-onic acid). The slower com-
ponents Avere simple proteins. In native synoAual fluid and native
Autreous body fluid the hyaluronic acid in all probability is in no
Avay combined Avith protein but exists solely in the form of salts
of the inorganic bases present. The glycoproteins precipitated
from neutral synoA'ial fluid or Autreous body fluid on addition
of acetic acid are to be regarded as purely artificial products.

^ There appears to be a definite need of a common name for the acid polysac-
charides occurring in animal tissues, With regard to the close relations of most
of these acids to mucins and mucoids the name “vniciiiic acids" (Mucinsauren,
acides muciniques) is proposed for these acids.

GUNKAR BLIX.

Neutral solutions of submaxillary mucin -svere, on electroplio-
xesis, found to contain two distinct components, botli proteins.
The main component, which was the faster migrating one,
amounted to about 90 % of the whole and was a glycoprotein
containing 27 % carbohydrate. The other component probably
ought to be regarded as a contaminant.

Pure water extracts from cartilage contain chondroitin sulphuric
acid exclusively in the free form. This confirms the old opinion
of Morxer that the chondroitin sulphuric acid in the cartilage
is partly present as alkali salt.

A chemical classification of the glycoproteins is suggested.

References.

Blix, G., Hoppe-Seyl. Z. 1936. 240. 43.

— , Skand. Arch. Physiol. 1938. SO. 46.

— , C. 0. Oldfelot and 0. KIarlberg, Hoppe-Seyl. Z. 1935. 234.111.
Hesselvik, H., Ibidem 1938. 2o4. 144.

— , Unpublished investigations 1940.

Kablberg, 0., Hoppe-Seyl. Z. 1936. 240. 55.

Levexe, P. a., Hexosamines and mucoproteins, London 1925.
Loxgsworth, L. G., and D. A. McIkxes, Chem. Rev. 1939. 24. 271,
esp. p. 283 ff.

ilEXER, K., Cold Spring Harb. Symp. Quant. Biol. 1938. 6. 91.
— , K., and J. W. PALiiEB, J. biol. Chem. 1934. JO/. 629.

— , and — , Ibidem 1936. JJ4. 690.

— , — , and E. M. Ssiyth, Ibidem 1937. JJ9. 501.

— , and E. M. Ssiyth, Ibidem 1937. J19. 507.

— , — , and M. H. Dawsox, Ibidem 1939. J2S. 319.

Morxer, C. Th., Skand. Arch. Physiol. 1889. I. 210.

Nilssox, I., Biochem. Z. 1936. 2So. 386.

ScHMiEnEBEBG, 0., Arch. exp. Path. Pharmak. 1891. 2S. 355.
SvExssox, H., Kolloidzschr. 1939. 87. 182.

Ibidem 1940. 90. 142.

Takahata, L., Hoppe-Seyl. Z. 1924. 136. 82.

TiSELitrs, A., Trans. Earaday Soc- 1937. 33. 524.

— , Kolloidzschr. 1938. 85. 129.

— , and E. A. Kabat, J. exp. Med. 1939. 69. 119, esp. p. 126.

From the Chemistry Department. Knrolinska Institutet. Stockholm,

Sweden.

A Colorimetric Metliod for tlie Deterininatioii
of Potassium in O.oi— O.i cc. of Blood Serum.^

K. A. J. WRETLIND.

OVith 1 figure in tbe text.)

In determining tlie potassium content of serum it sometimes
proves necessary to use such small quantities of serum as 0.1 c.c.
For that purpose a method has been worked out by the author
which is applicable to this quantity and also to a still smaller
amount, 0.01 c.c. of serum, e. g. in determining simultaneously
chlorine ions electrometrically. for which only 0.02 c.c. of un-
diluted serum is needed.

Of the methods proposed for the quantitative determination of
potassium in serum, those are the best ones in which potassium
is precipitated as potassium sodium cobaltinitrite and cobalt
(SoBEL and Kramer) or nitrite is determined colorimetrically.
The precipitation is preferably made in a buffered solution at
pH 5. G, according to Kramer and Tisdall, at which acidity the
serum proteins remain in solution. Tischer has shown that such
minute amounts of potassium as 1 y can be quantitatively pre-
cipitated as cobaltinitrite and the nitrite determined with accuracy
colorimetrically after adding Riegler’s naphthol reagent.

Taylor has worked out a method for 2 c.c. serum. He pre-
cipitates K in the deproteinized serum filtrate as cobaltinitrite
and determines the nitrite with Hosvay’s reagent (sulfanilic acid
and a-naphtylamine). Seeing that this reagent allows a deter-
mination of very small amounts of nitrite, it has successfully been
applied in this case. The sodiumcobaltinitrite-reagent of Kramer-

‘ Received for publication 30 March 1940.

K. A. J. WKETLIND.

Tisdall has been slightly modified. As recommended by Jacobs
and Hoffmait the precipitate was washed with distilled water and
70 per cent alcohol.

Principle.

The potassium is precipitated directly from the serum as
KoNaCo (N02 )o at; pH 5.6 — 6, by adding a buffered solution
of sodium cobaltinitritc. The cobaltinitrite precipitate is
dissolved in alkali. Colour is developed by means of Ilosvay’s
reagent.

Reagents.

1. A buffered sodium cobaltinitrite solution.

30 Gm. of HuaCo (NOals and 10.22 Gm. of NaAc arc dissolv-
ed in 210 c.c. of HiO. To this 0.47 c.c. of glacial acetic acid
are added. The reagent is kept in an ice-chest. It is to be
filtered before use. It can be stored for 1 — 2 months.

2. Distilled water saturated with octyl -alcohol.

This reagent is prepared by shaking 200 c.c. of water
with 2 c.c. of octyl-alcohol and then filtering through
hardened filter-paper,

3. 70 vol. per cent alcohol.

4. 2 per cent (0.5 N) Ha OH.

5. Ilosvay’s reagent.

0.5 Gm. of sulfanilic acid and 0.2 Gm. of a-naphtylamine
are dissolved in 320 c.c. of HAc (GO c.c. of glacial acetic
acid pro 1,000 c.c. of solution) The reagent is destroyed if
heated and it is to be kept in the dark. It must be colourless.

0.1 cc. of sernm.

Method.

In a small Pyrex centrifuge tube (6 X 50 mm) exactly 0.1 c.c.
of serum (not hemolysed) is pipetted off. To this 4 X 0.05 c.c.
of HajCo (N 02 )o-reagent are added, shaking thoroughly after each
portion of 0.05 c.c. being added. The tubes are left at room-
temperature for 2 hours. Then 0.2 c.c. are added of reagent 2.
After centrifuging about 20 minut-es at 3,000 r. p. m. the liquid
is carefully decanted, and the tube is allowed to stand upside
down on a filterpaper (5 min.) and dried with filterpaper. The
precipitate is washed by thoroughly suspending in another 0.5 c.c.
of reagent 2 and then centrifuged (5 — 10 min.). The liquid is

POTASSIUM IN BLOOD SERUM.

decanted, and tlie tube is dried upside down again. Then it is
washed twice in the same way with 0.5 c.c. of alcohol (70 %) each
time. After drying it the second time, the precipitate is dissolved
in 0.5 c.c. of NaOH (2 %) and then the tube is put in a boiling
water-bath for 10 minutes. Care is to be taken that no trace of
the precipitate remains on the walls of the tube. This is best
avoided by stirring with a glassrod. This procedure is the most
important one. The solution is poured into a 100 c.c. measuring-
flask, and the tube is washed 3 or 4 times with reagent 2. The
glass-rod is also washed, and the wash-water is poured into the
measuring flask, which is filled to the mark with distilled water,
h'or the determination of nitrite, exactly 5 c.c. hereof is transferred
to a 25 c.c. measuring-flask and 5 c.c. of Ilosvay’s reagent are
added. Also this measuring flask is filled to the mark with distilled
water. After 10 minutes the red colour is determined colorimetric-
ally in a Pulfrich-Photometer, with colourfilter 853 and 3 cm
cuvettes against a solution of 5 c.c. of Ilosvay’s reagent diluted
to 25 c.c. with distilled water.

Calculation.

If the determination is carefully carried out according to the
above method, the K of serum is calculated as follows.

mg% K = 55.0-E3“

= the extinction-coefficient
cm’s cuvettes with colourfilter S53.

I post’
I ante,

for 3

O.oi cc. of serum.

Serum is diluted in the proportion 1 : 10.

Of this 0.1 c.c. is pipetted into a small Pyrex-tube (6 x 50 mm).
To this 4 X 0.05 c.c. of HasCo (N02)s-reagent are added, shaking
it thoroughly after each portion being added. The tube is put
into an ice-chest (0° C) and is left there overnight (10 — 12 hours).
Then the operation is carried on according to the method for
0.1 c.c. of serum. After dissolving the precipitate the solution
is transferred directly to a 25 c.c. measuring flask. To this 5 c.c.
of Ilosvay’s reagent are added and the flask is filled to the mark
with distilled water. Colorimetric reading as above.

Calculation.

mg% E: = 27.50-E3«

K. A. J. WRETLIND.

4()

Controls.

At tlie first -SN'asliing with octyl-alcohol and distilled water,
there is no risk of the precipitate being dissolved, because there
is still Na 3 Co(N 02)6 present. The octyl-alcohol is used to make
the washing easier.

K 2 NaCo(N 02 )o is almost insoluble in 70 per cent alcohol. The
wash-water from the 3 rd. washing gives no nitrite-reaction with
Ilosvay’s reagent; the second, however, gives a noticeable reaction
There is no risk of precipitating IS^ajCofNOj)* with 70 per cent
alcohol.

When the precipitate is dissolved in NaOH the following reac-
tion takes place.

KsNaCofNO,)^ + 3NaOH = 2 KNOo -f 4 NaN 02 + Co(OH) 3 .

That nitrite can be determined quantitatively with Ilosvay’s
reagent according to the method above, is shown in table I,
where the formula c = • E (c = the amount of nitrite expressed

in y. = 5 .con 3 tant. E 3 '*® = extinction-coefficient) is proved to
be valid.

Table I.

y of NOj
presents as
NaNOj

Reading of
Photometer

K = -

Ej“

Deviation from
the average
valne ^

l.GO

67.60

9.4

-3.1

3.20

45.15

9.8

— 4.1

4.80

32.85

9.9

+ 2.1

6.40

22.0

9.7

±0

8.00

15.40

9.8

+ 1.0

9.60

10.60

9.8

+ 1.0

Average value 9.7.

Thus y NO 2 = 9.7 • E 3 ®®.

The K calculated as K 2 NaCo(N 02)6 will thus be

K = 2.75-E3'®.

That the precipitate has the formula KjNaCo (N 02)8 is shown
in table II. Here the K-level is calculated from the nitrite of
the precipitate, according to the above formula. As the value
obtained in this way, closely corresponds the known value of K,
the formula must be correct.

POTASSIUM IS BLOOD SEBUM. 47

TaWe II.

K-level in
solution

Time for the
precipitation

Dissolved
precipitate
used for the
nitrite

determination

Colonrfilter

s«

K-level

obtained

Differe.nce

1.31

12 h. O’C

1/1

32.5

1.34

+ 2.3

1.81

12 h. O’G

1/1

33.6

1.31

± 0

1.3)

12 h. 0° C

1/1

32.0

1.8G

+ 3.8

13.1

2h. 20’ C

1/20

57.0

13,6

+ 3.1

13.1

2h. 20° C

1/20

58.5

12.8

— 2.3

13.1

2 h. 20’ G

1/20

56.6

13.65

+ 4.2

In figure I tlie amount of potassium precipitated is plotted
against the time. It is shown that the concentration of potassium
influences upon the time necessary for complete precipitation.
jFrom these curves the time necessary for complete precipitation,
has been selected.

The importance of adding the sodium cobaltinitrite in small
portions has been pointed out by Eeamek and Tisdall.

In table III determinations have been made with 0.01 and 0.1
c.c. of serum with and without addition of potassium.

K. A. J. AVRETLIND,

Table UI.

Quantity of
Bcruin c,c.

K-level in
Bcrnm jng?»

Quantitj' of
K added
tng^

Total
quantity
of Ktig)*

Obtained
quantity
of Kmg %

■

O.Ol

27.0

13.0

40.0

39.1

- 2,25 1

O.Ol

19.2

16.0

3.5.2

34.6

-1.7 j

O.Ol

17.3

16.8

34.1

34.5

-f 1.2 I

0.1

25.8

12.6

38.3

37.8

- 1.3 I

0.1

26.5

12.2

38.7

39.5

+ 2.0 I

0.1

20.4

lO.O

30.4

30.2

-0.7 1

TaWe IV shows some values found in normal blood of man and

rabbit.

Table B'.

Quantity of

K-level mg% K

Bernm c.c*

Man

O.Ol

19.2

; 0.1

19.4

! 0.01

17.3

O.l

17.5

. O.Ol

20.4

; 9-1

20.3

Rabbit

! O.Ol

26.9 i

i 0.1

27.0 i

Summary.

A method has been described for determining the potassium
content in 0.01 — 0.1 c.c. of serum. The accuracy of the method
is satisfactory. Added amounts of potassium are quantitatively
recovered.

The author’s thanks are due to Dr. E. Jorpes at who’s sugges-
tion this investigation was taken up.

Literature.

Tischer, J., Biochem. Z. 1931. 238 . 148.

EjRAiiER, B. and E. F. Tisdale, J. Biol. Chem. 1921. 46 , 339.
SoBEL, A. E. and B. Kkasieb, J. Biol. Chem. 1933. 100 . 561.
Taylor, F. H. L., J. Biol. Chem. 1930. 87 . 27.

Ilosvay de N. Ilosva, M. L,, Bull. Soc. Chim. 1889. 2 . 347.
Jacobs, H. R. D. and Hoffma>', J. Biol. Chem. 1931. 93 . 685.

From the Physiological Institute of the University, Oslo, Norway.

A New Method for the Deterniination of Carbon
Monoxide in Blood.^

R. WENNESLAND.

(With 4 figures in the text.)

On working with experimental carbon monoxide poisoning the
following method for quantitative determination of carbon
monoxide in blood was evolved. The method requires no special
apparatus, is very easy to perform and the accuracy is great.

Principle.

The method depends upon the ability of carbon monoxide to
reduce palladium chloride after the following equation:

CO + PdCl. + HjO = Pd + CO, -h 2HC1.

This principle has been much employed especially for the detec-
tion and determination of carbon monoxide in air. A convenient
method for determination of carbon monoxide in blood depending
upon the same principle has been elaborated by Christman &
Eandall (1933). I have used their method as a starting point.
It includes two different steps:

1. The blood gases are extracted by reduced pressure after
admixture of acid ferricyanide, and passed over to a reaction cham-
ber containing a known solution of palladium chloride. The con-
nection between extraction chamber and reaction chamber is
closed.

2. After the reduction has taken place, excessive palladium
chloride is determined colorimetrically.

* Received for publication 10. JIaroh 1940.

4 — '{01323. Acta pliys. Scandinav. Vol. I.

R. \YESNESLAh’D.

On both steps I thought improvements could be made. At 1; It
seemed unnecessary to use reduced pressure for the extraction
of carbon monoxide. Regarding the latter, the diffusion rate de-
pends upon the ratio between the partial pressure of the carbon
monoxide in the liquid phase and in the gaseous phase. The
presence of pure air does not interfere with that ratio. The effect
of reduced pressure is merely a mechanical one caused by the
bubbKng. It seemed also unnecessary to transfer the extracted
gases to the reaction chamber and close the connection between
extraction chamber and reaction bulb. The gases can wander
by their own kinetic power, and the connection between the two

50cc. Erlenmeyer flask.

Blood acid Palladium chloride

mixture. solution.

Fig. 1. AnalyBis apparatus set up with blood acid mixture in the left Erlenmeyer
flask, palladium chloride solution in the right.

chambers may stand open as long as the blood reagent mixture
and the palladium chloride solution do not come in contact with
one another.

On this line of argument I made my apparatus, consisting of
two Erlenmeyer flasks fastened together at the open ends by
means of an airtight rubber cuff (Eig. 1).

Before they are connected, the blood reagent mixture is placed
in one of the flasks, a known solution of palladium chloride in the
other. The connected flasks are kept in a horisontal position,
carbon monoxide diffuses from the blood through the air to the
palladium chloride, where a corresponding quantity of the latter
is reduced to metallic palladium.

At 2: As to the determination of the excessive palladium
chloride I found the colorimetric method unsatisfactory. In the
quantities which I used, I could determine the palladium chloride

CARBOK MONOXIDE IN BLOOD.

colorimetTically mtli a standard doAdation of 0.0175 mg,
vrMcl) represents in carbon monoxide i 0.22 volume per cent,
supposing that 1 cc. of blood has been used. With photometric
determination I got only insignificantly better results. I worked
out a titrimetric method depending upon the following reaction:

PdCh -f 2KI = Pdl- + 2KC1.

The palladium chloride is precipitated with a known quantity
of potassium iodide and the excessive potassium iodide is deter-
mined iodometrically after the principle of Dupre-Winkler:
Iodide is oxidised to iodatc by means of chlorine water (Von
Weszelsky 1900, Winkler 1916) or better bromine water (Bu-
GARSZKY & Horvath 1909,) after the following equation:

KI + 6C1 -f SH.O = KIO, -k 6HC1.

Superfluous chlorine is removed by boiling, or if bromine has been
used, it can also be removed by the addition of phenol in excess
(Schulek k Stasiak 1928,), h^rom the potassium iodate formed,
iodine is released by adding excess of potassium iodide in strong
acid solution according to the equation:

KlOa + SKI + 3H.SO* = 31* + ZKSO, -f 3HiO.

With this method I could determine excessive palladium
chloride with a standard deviation of ± 0.00137 mg. which repre-
sents in carbon monoxide, supposing 1 cc. of blood has been used,
± 0.017 vol. per cent, that is 13 times the precision of the colori-
metric method.

Precipitation with potassium iodide for the determination of
palladium chloride is an old method, but I cannot find that anyone
has applied the principle in the above mentioned manner before.
IsHiSAKA (1937) has published a method for the determination of
carbon monoxide in air based upon the palladium chloride reaction,
where "the concentration of PdCl* is determined by titrating
with KI solution,”

Eeagents.

10 per cent sulfuric acid for the transformation of carbon mon-
oxide hemoglobin into sulfuric acid hematin. Christman &
Eandall used ferricyanide after Haldane (1897 — 98) mixed

R. WENNESLAND.

witli lactic acid, to free tlie carbon monoxide, I found however
that during the interval until the reduction of palladium chloride
is maximal (3 — 4 hours), a certain amount of the acid ferricyanide
is converted into a volatile substance, probably hydrocyanic acid,
which reacts with the palladium chloride and gives higher values
for the carbon monoxide than actually present. E. g. a blood
sample containing 5.30 vol. per cent of carbon monoxide gave
values of 8.48 vol. per cent after 4 hours, 9.05 vol. per cent after
13 hours, when treated with acid ferricyanide.

1/400 N potassium iodide solution containing 1 per cent of
aluminium sulfate. — 0.415 grams of potassium iodide p. a. are
dissolved in 1 bter of distilled water containing 10 grams of alu-
minium sulfate. The solution is stored in a dark bottle. The
aluminium sulfate serves to precipitate colloid metallic palladium
and palladium iodide (Christman & Eandael).

Saturated bromine water produced by shaking pure bromine
with distilled water and stored in a bttle bottle with glass stopper.

1/200 N sodium thiosulfate solution. — 1.241 grams of
NajSaOa'SHaO are dissolved in 1 liter distilled and boded water
containing 10 grams of amylic alcohol. With amyhc alcohol as a
conserving fluid the thiosulfate solution keeps its concentration
very well for months,

1/200 N potassium iodate solution for the adjustment of the
thiosulfate, — 0.3567 grams of potassium iodate p. a. (Merck)
are dissolved in 2 bters of distilled water and kept in a glass-
stoppered bottle in a cool place.

1 per cent starch solution. One gram of soluble starch is dis-
solved in 100 cc. of boiling water saturated with sodium chloride.
I have also added 1 per cent of amylic alcohol for conservation
instead of sodium chloride.

2.5 per cent phenol in 2.5 per cent sulfuric acid. A 5 per cent
solution of phenol in water is mixed with equal parts of 5 per cent
sulfuric acid.

1/100 hr palladium chloride solution. — 0.8881 grams of PdCh
p. a. (Merck) are dissolved in 100 cc. of 1/10 N boiling hydro-
chloric acid. After cooling to room temperature the solution is
diluted to 1 Hter. Kept in small bottles with glass stoppers the
reagent keeps its concentration very well for long periods (Table
I). In accurate working however, blank analyses will be made, and
it is of minor importance how the palladium chloride solution is
stored and how it keeps its concentration.

CARBON MONOXIDE IN BLOOD.

Table 1.

The DnrahUiiy of the Palladium Chloride Solutions.

Cnlcnlatcd concentration of the
various solutions

Age of so-
lutions

Concenfratio
tions deter
present

Stored in cle.ir
bottles

1 of the sola-
mined after
method

Stored in darh
bottles

Daj’S

Kormalit^'

Normality

Solution 1; 0.1T6 gin. of PdCI, dis-

0.0097U

solved in 200 cc. of dilnt. IlCl

0.00977

N = 0.00985

0.00977

1 Solntion 2: The E.imc concentration

0.00979

0.00P77

0.00978

■ .

0.00979

' Solution 3: 0.,S&0 gm. of PdCl. dis-

0.01961

0.01951

solved in 200 cc. of dilnt. HCl

0.01944

0.01943

N = 0.01970

0.01949

0.01948 i

0.0194S

0.01949 !

0.01961

0.01941 j

0.01963

0.01966

0.01946

Apparatus and Equipment.

Handling of tlio blood: To prevent the clotting of the blood
heparin, potassium oxalate or sodium citrate may be used, as
experience has shown that none of these substances interferes with
the analysis. I consequently add 0.15 per cent of dry neutral po-
tassium oxalate.

The blood may be handled and stored anaerobically after the
directions in Peters & Van Slyke (1932). More convenient
however is the use of syringes with glass piston (Barcropx &
Haldane 1902, Scholandeb 1938). The blood sample is drawn
into an airfree syringe containing the potassium oxalate or another
anticoagulant. The syringe is provided with a smaU rubber tip,
which Sin be closed by a short nail. Syringes where the tip is

R. WRNNESLAND.

placed excentrically, are especially convenient when small air
bubbles are to be removed. The sample can be stored about 2
weeks without changing its content of carbon monoxide when kept
at a temperature of 3 — 5° C. When kept at room temperature,
the evolution of putrefactive gases woll some times interfere with
the analysis after 2 — 3 days. The presence of interfering substances
is detected quite easily however (Sec: Remarks on details).
Apparatus: The form of the apparatus after the principle two
chambers in open connection with one another, may naturally

Rubber ribbon.

Cork disc.

Wood or metal bar.

NN.CsJ.'

Fig. 2. A. One of the prepared cork discs with rubber ribbon. B. Part of the
rotary mechanism with analysis apparatuses placed in one of the cork discs.

be varied in different ways. I have found two 50 cc, Erlenmeyer
flasks to be the best, fastened together at the open ends as pre-
viously mentioned, by means of a rubber cuff of suitable calibre.

I have found it practical to make a rotary mechanism for the
flasks (Fig. 2), A long wood or metal bar penetrates the centre
of a series of cork discs placed at a distance of 15 cm. from each
other. The corks have a diameter of 7 cm. and a thickness of 3 cm.
Four small nicks are cut in the circumference of the corks, where
the neck of the flasks can rest firmly. The flasks are kept in place
by a 2 — 3 cm. broad rubber ribbon, which is tied once around
the cork with the flasks and fastened on a nail. The bar rests
horisontaUy on two bearings and is rotated by a motor. 50 — 60
times a minute. In series analyses especially the rotary mechan-

CAKBOK MONOXIDE IK BLOOD.

ism is practical. By making the bar long enough and using large
corks, there is room for a large number of flasks.

Of adjusted pipettes a 1 cc. pipette is needed for the blood,
further a 2 cc. a 5 cc. and a 10 cc. one. All pipettes are adjusted
with water for delivery after the instructions in Peters & Van
Slyke (1932).

For the final titration a 10 cc. burette with 1/100 cc. gradations
is used. With a fine glass tip droplets of 0.01 — 0.02 cc. can be ob-
tained. The burette is adjusted with water.

Procedure.

Both of the flasks must be clean and dry. One of the flasks is
prepared with the rubber cuff on its neck and 2 cc. of palladium
chloride solution in it. The other flask contains 3 cc. of distilled
water. 1 cc. of blood is delivered slowly under the water. The
flask is quickly shaken to mix the blood and the water, cc. of
10 per cent sulfuric acid is dropped into the blood water mixture,
and as quickly as possible the flasks are connected in a horisontal
position without letting any of the blood acid mixture come in
contact with the palladium chloride.

The flasks are placed in the rotary mechanism and rotated for
3 — i hours, while the reduction of palladium chloride takes place.
Then they are disconnected and the excessive palladium chloride
is determined in the following manner; To the flask containing a
mixture of metallic palladium and palladium chloride, 10 cc. of
the K1 — AlofSOila mixture is added, and the flask is shaken
vigorously. The palladium iodide and metallic palladium will
form a fine dark brown precipitate, which is separated from the
solution by filtering through a dry filter into a dry flask. From the
filtrate an aliquot portion is taken for the determination of ex-
cessive potassium iodide. I have found 5 cc. out of a total fluid
quantity of 12 cc. suitable. Into these saturated bromine water is
dropped until the solution shows a bright yellow colour. 3 — i
drops will generally suffice. The flask is shaken and after 1 — 2
minutes the excess of bromine is removed and the solution made
acid by adding 2 cc. of the phenol sulfuric acid mixture and shak-
ing. One minute or so later a crystal or two of potassium iodide
p. a. is added and after shaking, the !• formed is titrated with the
adjusted 1/200 N sodium thiosulfate solution. The titration must
be done slowlj'. VTien the fluid is nearly decoloured. Vs cc. of the

K. WBNNESLAND.

starch, solution is added and the titration goes on until the blue
colour disappears. With good starch solution the change w-ill
occur with — 1 droplet, that is 0.01 — 0.02 cc. of 1/200 N sodium
thiosulfate.

For accurate results blank analyses ought to be made to correct
for small quantities of carbon monoxide in the laboratory air which
is enclosed in the flasks. The blank analyses are made as the blood
analyses regarding all reagents except that the blood is omitted.

Modifications in the procedure: "When an analysis has to be done
only occasionally, it will not be profitable to make a rotary
mechanism. The results however will be as good when the flasks
are left lying still at room temperature. In that case the reaction
time must be 14 — 18 hours to reach maximum values. The blank
analyses may be omitted and be substituted by a titration of 2 cc,
of the palladium chloride solution after the present method now
and then. The error introduced thereby is generally negligible.

Calculation.

The concentration of the palladium chloride is calculated after
the equations:

PdCh + 2KI = Pdl* + 2KC1.

KI 6Br -}- SHjO = KIO3 + 6HBr.

KIO3 -f 5KI + 3H,S04 = 31. + 3K3SO4 -f 3H,0.

The formula for the calculation of the concentration of the
palladium chloride thus will be:

where a represents the cubic centimeters of thiosulfate required
to titrate ad modum Dupre-Winkler the quantity of KI added
to the palladium chloride solution, b the thiosulfate required to
titrate excessive KI, when the Pdl. precipitated has been removed
by filtering. X represents the normality of the thiosulfate adjusted
with 1/200 N KIO3, and q the aUquot portion taken, generally
5/12.

The carbon monoxide is equivalent to the quantity of palla-
dium chloride reduced, which can be expressed as the difference

CAR’JOX MONOXIDE IN BLOOD. oT

between original PdCl. (or better PdCl- of the blank analysis) and
excessive PdCl* after the reaction with carbon monoxide.

As ai = a; (generally one will add as much KI to the blank as
to the blood analysis), the expression is reduced to:

CO =

b» — bi
6q

where b^ is the cubic centimeters of thiosulfate required to titrate
the blank, b™ is the thiosulfate required for the blood analysis.
X and q have the same meaning as previously.

From the equation:

PdCl, + CO + H,0 = CO. + 2HC1 + Pd

22.4

one will see that 1 cc. of N PdCl. ~ — ^ cc. of CO at 0° C. and

760 ram. Hg. — To express the carbon monoxide as volume per

100

cent, one will have to multiply the expression above by

where v is the volume of the blood sample. The complete expres-
sion for CO in volume per cent of the blood is:

(b. — b,) X 1120

6 vq

With amylic alcohol added to the thiosulfate X will remain
constant, and the expression above can be abbreviated to:

CO = (b. — bi) f volume per cent,

where f is constant = -7; as long as the same concentration

o V q

of thiosulfate is used, the same volume of blood and the same
aliquot portion are taken,

E.xaraple: Analysis of 1 cc. of blood has given a terminal excess
of KI = 9.45 cc. of 1/205 N sodium thiosulfate.

The blank analysis gave excessive KI = 2.85 cc. of thiosulfate.
The aliquot portion taken was 5/12.

K. WKNNESLANI).

f)8

(9,45 — 2.85) f vol. per cent, where
12 X 1120

f =- rr= = 2.185

CO = G.GO X 2.185 = 14.42 vol. per cent in the blood. (For
correction see below under: Tle.sult.s.)

Rcmnrks on Details.

The dilution of the blood with water before addition of sulfuric
acid is important. If not well mixed with water, the blood some
times will form a clot when sulfuric acid is added, and the analysis
can be spoiled. Especially when the flasks arc left lying still, the
blood acid mixture often will form a gel. This docs not however
interfere with the analysis. Sufficient acid must be used to con-
vert all hemoglobin into sulfuric acid hematin. ’/* — ’/s cc. of 10
per cent H.SO4 will be abundant for 1 cc. of blood. If all the
hemoglobin is not converted into sulfuric acid hematin, the mix-
ture will keep a red tinge, different from the dirty brown colour
of the acid hematin. The reaction goes on in the same way even
if sulfuric acid is not added to the blood, but much more slowly
and the results vary more.

The effect of shaking on the reaction time was examined. The
results are presented graphically in fig. 3.

■\\Tien the flasks are rotated, maximum values are attained af-
ter 3 — 4 hours reaction, ^yter about 6 hours a slight decrease
takes place and goes on as long as I have observed it.

I^Tien the flasks are Ijnng still, maximum values are attained
after 14 — 18 hours, whereafter a similar decrease takes place. I
tried to shorten the reaction time by warming the flasks and the
reagents, but with no essential result.

Winkler (1934) in a publication on a method for determination
of carbon monoxide in air, depending upon the palladium chloride
reaction, has made exactly the same observation as to the corre-
lation between reaction time and maximum values when the flasks
are shaken.

As to the slight decrease in the values for carbon monoxide as
the length of reaction time increases, the cause is perhaps as indi-
cated by Christman, Block & Schultz (1937) that the oxygen
in the air interferes with the reaction of palladium chloride with
carbon monoxide.

CARBON MONOXIDE IN BLOOD.

Fig. 3. Annlypcs of Blood for Cnrhon Monoxide.

Curves showing the correlation between annUnis values and reaction time,
when the flasLs have been rotated and when they have lain

fitill — • —. All curves are empirical. The analysis values arc uncorrected.

The thin horizontal dotted lines represent the corresponding values

determinetl after the method of .Sendroy & Liu.

As the figure shows, tlie maximum values of carbon monoxide
after the described method lie about 4 per cent of the values
concerned lower than the "real” carbon monoxide content deter-
mined after the method of Sendboy & Liu. Tliis error is rather
constant for all quantities, as will be seen later, and is naturally-
explained as a consequence of the interfering process mentioned
above.

The palladium chloride solution must not be too acid in view
of the later oxidation of excessive potassium iodide with bromine,
which according to all authors is appreciably disturbed if the pH
is too small. (E. g. Eeith (1929) whose suggestions I have fol-
lowed.) In the filtrate I found the pH about 5.4 before the addition
of bromine water. For the oxidation of KI to KIOj I have tried
both chlorine CWinkler) and bromine (Bugarszky & Hor-
vath) with equal results. Bromine is preferable because it is more
conveniently handled and more easily driven out.

R. ^VRNKESLAND.

I have also tried other methods for determination of excessive
KI, such as direct titration with potassium iodate in strong liv-
drochloric acid solution, (Andrews, after Kolthofp 1931,) and
the “iodo-cyane method” of Lakg (Kolthoff 1931). Titration
after the principle of Dupre-Winkeer has however giveii me the
best results, probably because the titration error is divided by 6,
as will be seen from the reaction equation.

The Erlenmeyer flask containing the palladium chloride solu-
tion must be scrupulously cleaned and dried. It can be washed
with dilute ammonia and rinsed well with warm water. It is prac-
tical always to use the same flasks for the palladium chloride and
others for the blood. In the flasks with the blood acid mixture
traces of a colloid clot Avill often adhere to the walls and are diffi-
cult to remove. This does not interfere with the analysis, when the
same flasks are used for the blood, but some times when I happened
to take such a flask for palladium chloride, the solution got brown-
ish, cloudy and gave too low values for the carbon monoxide.
Other impurities in the analysis, ns when I originally mixed the
blood with acid ferricyanide, and sometimes when I examined
very decomposed blood, were always detected because they dis-
colored and clouded the palladium chloride solution. 1 take it as
an indication that the reaction is without impurities, when the
PdCl» solution keeps a bright yellow colour and the metallic palla-
dium after some time floats on the surface like a mirror.

Regarding the preeijiitation of palladium iodide. Beamish &
Dale (1938) have stated that at least ten times the calculated
amount of potassium iodide may be added without danger of
palladium loss. Tidicn using reagents of the same concentrations
as I do. the ratio IvI/PdCh = 10/8 when no carbon monoxide is
present, and 10/1 when the carbon monoxide content is 19.G
vol. per cent. Up to the latter content the requirements of Bf.a-
.Mi.SH & Dale arc thus fullfillcd. If larger quantities arc to be
e.xamincd, a somewhat stronger solution of palladium chloride
can be employed, or the blood sample may be reduced to ‘/s
These precautions are advisable also because 2 cc. of l/lOO R
palladium chloride arc only capable of oxidizing a theoretical
ma.ximum quantity of 22.4 vol. per cent of carbon monoxide m
1 cc. of blood.

For filtering I u.se white ribbon filter paper, Schlkichkr &
ScHULE No. 589=. Any filler which does not allow particles of
Pd or Pdl. to pas.s, i.s apjilicable. If traces of Pd or Pdl; paes the

GAKBOK MONOXIDE IN BLOOD.

filter, they Avill be detected at the final titration with sodium thio-
sulfate. The change of colour will be indistinct and the final solu-
tion which normally is water clear, will show a red brownish tinge.
The same filter can be used for all double analyses.

It is of importance that sufficient bromine is added, but not
too mucli. If too much, it will form with phenol a cloudy grey-
yellow precipitate and this interferes witli the analysis. After
the addition of phenol sulfuric acid mi.xture the solution ought
to remain clear (Gloss 1931.) In most of my analyses excessive
bromine was removed by boiling after Eeith’s instructions (Reith
1929). Removal with phenol is more convenient and as reliable.

Results.

As control I have used the method of Sendkoy & Liu for
determination of carbon monoxide in blood (Sendroy & Liu
1930, Peters & Van Slyke 1932) a manometric method which
I have modified a little, introducing a syringe absorption pipette
instead of the modified Hempel pipette of Van Slyke-Hiller.
(Author in press.)

To test the reliability of my own method I produced series of
blood samples with various carbon monoxide content. Generally
I used defibrinated ox- or calf blood which I obtained from the
municipal slaughter-house, some times directly from the animal,
in the latter cases prevented from coagulation by means of
sodium oxalate. The blood was saturated with carbon monoxide
and its content determined after Sendroy Sc Liu’s method.

The saturated blood was mixed with unsaturated, whose quite
small content of carbon monoxide had been determined in the
same way. Later I found it more convenient and as exact to dilute
the saturated blood with distilled water in various proportions,
hly main series is made in the latter way.

The carbon monoxide content of the various blood water mix-
tures was determined in two ways:

1. It was calculated from the values of the saturated blood
and the known dilution grade (“calculated values”).

2. It was determined directly for all dilution steps after the
method of Sendroy & Liu (“determined values”).

The various dilution steps were: 1/10 — 2/10 — 3/10 — 4/10 —
5/10 — 6/10 — 7/10 — 8/10 — 9/10 — and undiluted.

R. WEN^'ESLA^•D.

Table II.

Analt/Fcs of Blood for Carbon Monoxide.

Present method. Renction time 4 hours. 1 cc. of blood has been used
for nil analyses. CO content expressed ns vol. per eent of blood.

Xnmber of
observations

Vol. per cent
of blood
a

Standard

deviation

Coefficient of
variation

-100

Bclation
between
j determined
valncsj (table
III) and a

V.O

l.r.7

0.032

2.0

1.038

VlO

2.')

3.10

0.050

1.8

1.030

Vio

4.75

0.048

1.0

1.038

V>0

6.10

0.072

1.1

1.042

V.O

8.04

0.070

0.8

1.080

V.O

10.35

0.050

0.5

1.046

Vio

2.3

12.00

0.1 04

0.9

1.082

V.O

2-1

14.04

0.0G2

0.4

1.020

V.O

15.82

0.111

0.7

1.048

>V.O

17.77

0.115

O.O

1.088

Table III.

AnaJy.’ies of Blood for Carbon Monoxide,

Present method compared with the method of Skndboy & LlU. 1 cc.
of blood used for all nnnlyses. CO content expressed ns vol. per cent

of blood.

Present method i Sendroj’ k Lin’s method

Jilntion

prade

, Vol. per
rent mnlti-
‘ plied bv
1.04 ’

! Xnmbcr of
j observa-
1 tions

»I)ctennined valnes»

»Cnlcnlatcd values » j

Vol. per
cent

Number
of obs

Vol. per ;
cent j

Number ;
of obs 1

Vi*

1.6.3

1.03

}

1.04 1

4 i

1 3.22

: 25

3.21

3.27 1

4 1

’/i.

! 4.94

{

1 29

4.93

4.92 1

3 1

v»

’ 6.05

: 53

6.07

6.50 i

3 !

*/tJ

8.99

8.95

8,96 1

! 10.76

; 20

10.83

10.78 ;

3 1

12.54

12.44

12..57 1

3 i

; 14.C0

14.40

14.87 i

3 i

; 16.45

• Of)

16,50 j 7

16.01 ;

1! ;

1 18.48

: 20

18.45

CARBON MONOXIDE IN BLOOD.

The results are presented in table II and III. As ivill be seen
in table II, the coefficient of variation decreases with increasing
carbon monoxide content, until from a content of about 8.5 vol,
per cent and higher it reaches an average of about 0,6 per cent of
the analysis value.

The coefficient of variation is the standard deviation in per cent
of the analysis value. The standard deviation is found after the
formula

where Oj represents the observations and n the number of the
observations.

As preidously mentioned the values of carbon monoxide after
the present method lie a little lower than the “real” values deter-
mined after the method of Sbndroy & Liu, and therefore have to
be corrected. The relation between the “determined values”
and the corresponding values found after the present method was
calculated for the various dilution steps. The results are seen in
the last column of table II. The relation amounts to an average
of 1,0375 and seems to be constant for all quantities of carbon
monoxide which are represented in the table, when a reaction time
of 4 hours is used.

In a supplementary series of analyses consisting of 113 observa-
tions after the present method compared to 58 observations after
the method of Sbndroy «&; Liu, ivhere the carbon monoxide con-
tent ranged from 3.54 vol. per cent to 18.42, the relation amounted
to an average of 1.043.

The values found after the present method therefore have to
be corrected by multipl}dng them by 1.04 to reach the values
obtained after Sendroy & Liu. In table III the values of my
main series have been corrected and are compared to the so-called
“determined values” and "calculated values.”

I tried to get a verification of my results in a more direct way
by analysing a mixture of air and carbon monoxide under approxi-
mately the same circumstances as the blood analyses. Carbon
monoxide gas was prepared in the usual way by warming a mix-
ture of anhydrous formic and sulfuric acids. The gas was washed
by bubbling it repeatedly through alkaline pyrogallate and ana-
lysed in an apparatus resembling the volum^fic apparatus Of

31. \V3:NJ.'ESLAND.

VA^• SiA'KE, but of only 10 cc. volume. The apparatus was filled
with the gas to the 10 cc. mark, carbon monoxide was absorbed
by Winkler’s cuprous chloride solution (prepared after Peters
Sc Van Slyke 1932), and the volume of the rest gases (a little N.
and water vapour) read on the calibrated upper stem of the
apparatus. 10 cc. of the gas analysed were mixed with 1000 cc.
of air and 10 cc. of the mixture were introduced into an Erleu-
meyer flask of 100 cc. volume containing 2 cc. of the usual palla-
dium chloride solution. The introduction was performed in the
following way: AYhcn the palladium chloride had been placed into
the Erlenmcycr flasks, they were tightly stoppered by rubber

jgfjj

—

■

[i!

—

iHln

■

-J

Inin

■

o .

-idjK

b.1.

■

—I

□

■

L 1

■

;■

■

1 !

■

.Mm

Li ^

20-

■

—Lilli

Uf!

sJ_

Fie. -I. Annlj'scs of 10 cc. of n Carbon Monoxide — Air Mixture. Empirical curve
showing correlation between analysis vnluc.s and reaction time.

stoppers penetrated by a glass tube with a stop cock. A little of
the air in the flasks was sucked out by means of a water pump,
sufficient to allow 10 cc. of the carbon monoxide — air mixture
to enter. The latter were introduced from the analysis apparatus
mentioned above. The tube and its stop cock were sealed with
water. A quantity of carbon monoxide corresponding to the con-
tent of 1 cc. of blood half-saturated ■with the gas was in this ,way
introduced into the fla.sks, which were rotated in the rotary
mechanism, and the rest of the analysis was performed as described
for the blood analy.ses. Experiments showed that ma.xiraum
values were attained after a reaction time of l‘/i — 2 hours (Fig- 4),
iiulicating that part of the longer reaction time of the blood
analy.ses is due to the slower diffusion rate when the ga.s 1ms to
pa.ss out of the blood reagent mixture through the air, before it
••liter.'; the palladium chloride .solution. The values of the analy.se.s

CARBON MONOXIDE IN BLOOD.

were corrected for temperature and barometric pressure and com-
pared with the “calculated” concentration of tlie carbon monoxide
in tbe gas mixture. Blanks were made as for the blood analyses.
The results are seen in table IV.

Table lY.

Analyses of 10 cc. of a Carbon Monoxide — Air Mixture.
Present method. Reaction time l.B — 2 hours.

Calculated CO
content.

Vol. per cent.

Determined CO
content.

Vol. per cent.

Relation between
calculated and
determined values.

Sample 1.

0.956

0.900

1.062

0.966

0.893

1.070

0.890

1.074

Sample 2:

0.965

0.912

1.047

New mixture.

0.914

1.046

CO the same.

0.912

1.047

0.909

1.061

0.908

1.052

Sample 3;

0.964

0.904

1.055

New mixture.

0.953

0.904

1.065

CO the same.

0.89G

1.064

Sample 4:

0.979

0.922

1.062

New CO produced

0.979

0.922

1.062

and mixed.

0.922

1.062

0.922

1.062

The relation between the "calculated values" and the analysis
values are seen to be a little greater than the blood analyses indi-
cate, viz. 1.058 per cent on the average. The mixing of carbon
monoxide with air and the transfer of the gas mixture to the ana-
lysis apparatus however, include sources of error through small
leakages, aU of them tending to increase the difference. I am
inclined to believe the difference of 4 per cent obtained from the
blood analyses to be the most correct one.

It is convenient to include the correction in the factor f (com-
pare the chapter; Calculation), by multiplying it by 1.04. The
5 — i0132S. Acta phys. Scandinav. VoLI.

R. -VVEKKESLAND.

fiG

Table Y.

The Correded Fador F Deiermincd for 1 cc. of Blood, an Aliquot
Portion of and Values of p from 4.90 cc. to 5.20 cc.

(t = the number of cc. of sodium thiosulfate solution required to titrate 5 cc. of
1200 N standard potassium iodato solution.)

y found
cc.

The conected
factor F.

y found
cc.

The corrected
factor F.

4.&0

2.877

5.00

2.802

4.91

2.372

5.07

2.297

' 4.92

2.867

5.08

2.298

i 4.93

2.803

5.09

2.288

; 4.91

2 858

5.10

2.284

i 4.95

2.353

5.11

2.279

j 4.90

2,348

5.12

2.276

j 4.97

2.344

5.18

2,271

1 4.98

2.389

5.14

2.200

j 4.99

2.334

5.16

2.202

1 5.00

2.830

5.10

2.257

1 5.01

2.325

5.17

2.263

i 5.02

2.820

5.18

2.249

' 5.03

2,310

5.19

2.244

1 5.0<

2.311

5.20

2.240

i 5.05

2.807

corrected factor is called F. In tabic V I have calculated F for
different concentrations of sodium thiosulfate solutions of about
1/200 N, u.Ring the formula:

1120 X

F = L04 1 = 1.04-- .

G vg

Supposing that 1 cc. of blood has been taken (v = 1,) and an
aliquot portion of 5/12 (q = 5/12,) the concentration of the sodium
thiosulfate (X) remains the only variable upon which the factor
depends. X is found by titrating 5 cc. of a 1/200 N standard solution
of potassium iodate. Calling the number of cubic centiraeter.H of
.■KKlium thiosulfate solution required in this titration y, X is found

from the equation: It is however most practical to

substitute X by y in the table, so that the factor F can be read

CABBON MONOXIDE IN BLOOD.

directly wlien 6 cc. of the potassium iodate have been titrated.
The carbon monoxide content is then found from tbe formula:

CO = (b, — bi) F.

If concentrations of sodium thiosulfate are employed which He
outside the quantities of the table, blood samples other than 1 cc.
and ahquot portions different from 6/12 are used, F must be deter-
mined from the formula above.

Table TL

Analyses of Blood with Small Contents of Carbon Monoxide.

’resent

m e t h 0

>C.a]cnIated

Cc. ana-
lysed

Vol. per cent
mult, by 1.04.

Standard

dev.

Number
of obs.

values.*
Vol. per cent.

Sample 1.

0.184

0.019

} 0.20G

Dilut. 1/100.

0.198

O.OIB

Sample 2.

0.093

0.030

\ 0.103

Dilut. 1/200.

O.094

0.016

Sample 3.

0.198

O.OIC

i 0.194

Dilut. 1/100.

0.193

O.ooy

Sample 4.

0.119

0.022

> 0.097

Dilut. 1/200.

0.114

0.013

In samples 1 and 2 the CO saturated blood was diluted vitb vater, and the
reaction time varied from 18 to 40 hours. In stimples 3 and 4 the saturated
blood vas diluted with fresh blood and the reaction time was constant: 3 hours.

I have finally analysed blood with a very small carbon monoxide
content (Table VI). The samples were prepared as described for
my main series by diluting CO saturated blood after having ana-
lysed it. Samples 1 and 2 were diluted 1/100 resp. 1/200 with
distilled water and blanks were made as usual. The reaction time
for the analyses of both of them varied much, from 18 to 40 hours,
because they were examined before I had detected the correlation
between time and analysis value. The long reaction time employed
is probably the reason that the usual correction of 1.04 is too small
for samples 1 and 2.

R. WEKXRSLAXD.

Samples 3 and 4 were analysed after I had prepared my main
scries and developed the technique. Though preliminary experi-
ments had shown that it was indifferent for the exactness of my
method whether blood diluted with water or mixed with CO-frec
blood was examined, I thought that it might perhaps play a part
at very small CO contents. Possibly the larger content of colloids
in a sample mixed with blood could retain minute quantities of
carbon monoxide, which generally had no influence on the precision
and accuracy of the method but might be detectable in analyses
of verv' small values. I therefore mixed the saturated blood
in samples 3 and 4 with fresh blood, and used corresponding quan-
tities of the latter for blanks to exclude errors through the presence
of minute quantities of carbon mono.xide in it. A constant reaction
time of 3 hours was used. In all analyses in table VI the mean
values of the present method are compared to the "calculated
values” only, determined by di\'iding the mean value of 3 qnalyses
of the saturated blood by 100 and 200 respectively.

As will be seen, the precision of the analyses of samples 3 and 4
is e.xcellcnt. The analyses of samples 1 and 2 are not so precise,
but as mentioned above, they were performed before the tech-
nique was fully worked out. E. g. the variations in the reaction
time used will tend to give a greater scatter of the results. At the
small values employed in the last experiments the precision of the
blanks will be only insignificantly better than that of the analyses.
Increasing the number of the analyses wdll therefore have no in-
fluence on the accuracy of the determination, unless the number
of the blanks is increased correspondingly. The analyses of sam-
ples 3 and 4 arc based upon the mean value of 7 blanks. The results
indicate that the method is well fitted for micro-analyses.

If one is only interested in the detection of carbon monoxide
in the blood, the present method is a very simple one for the pur-
pose. The long reaction time can be spared, as it is of no interest
for the detection that maximum reduction takes place. The titra-
tion is omitted and blanl^ are generally unnccessar}'. The two
Erknmeyer flasks are prepared as previously described with blood
sulfuric acid mixture in one, palladium chloride solution in the
other. V’ith blood samples of 1 cc. the smallest quantities of
carbon mono.xide which are distinctly observable arc about O.a
vol. per cent. The reduction was clearly seen after 20 — 25 minutes.
2 vol. per cent of CO gave dbtinct reduction after 10—12 minutes
and with 5 vol. per cent and larger it took about 5 minutes to get

CARBON MONOXIDE IN BLOOD.

a distinct mirror of metallic palladium on the surface of the solu'
tion. The flasks were rotated by rolling them on a table. If they
were left lying still, distinct reduction took place only 2—3 minutes
later. In a few cases, when the flasks were a little greasy on the
inside so that the palladium chloride solution did not moisten the
walls, no metallic palladium was precipitated. Titration in the
usual way showed that the solution contained reduced palladium
corresponding to a CO content of 2.70 vol. per cent. Controls
showed that it made no difference for the titration whether the
palladium was precipitated at once or kept in solution imtil it
was flocculated by the aluminium sulfate in the potassium iodide
solution. Values of carbon monoxide from 0.5 to 6 — 7 vol. per cent
can be roughly estimated with practice from the solidity of the
palladium mirror. From about 7 vol. per cent and larger it was
difficult to see any difference. The quantitative determination
however is so simple and convenient that it is preferable in most
cases.

Summary.

A new method for the determination of carbon monoxide in
blood is described. The blood is brought into an Erlenmeyer
flask, mixed with water and its carbon monoxide hemoglobin
is converted into acid hematin by addition of sulfuric acid. The
carbon monoxide liberated passes into another Erlenmeyer flask,
which has been connected with the first one by an airtight rubber
cuff and contains a known solution of palladium chloride. A cor-
responding quantity of the latter is reduced to metallic palladium.
Excessive palladium chloride is precipitated by a known quantity
of potassium iodide, and the excess of the latter is titrated after
the principle of Dupre-Winkler. The method is well fitted for
micro-analysis.

References.

Barcroft, J. and J. S. Haldane: J. Physiol. 1902, 28, 232.
Beamish, F. E. and J. Dale: Ind. eng. Chem. — Analyt. Ed. 1938, 10,
697.

Bugarszkv, S. and B. Horvath: Z. anorg. Chem. 1909, 63, 184.
Christman, A. A., W. D. Block and J. Schultz: Ind. eng. Chem. —
Analyt. Ed. 1937, 9, 153.

Christsian, a. a. and E. L. Randall: J. biol. Chem. 1933, 102, 595.
Closs, C.; tiber das Vorkommen des Jods im Meer und in Meeres-
organismen. Oslo 1931.

R. WENKRSLAh'D.

Halpane, J.: J. Physiol. 1897— 189S, 22, 29S.

IsnisAKA, 0.: J. Pharm. Soc. Japan. 1937, 57, 1007. (Quot<!d from
Clicm. Ahstr. 193S, 32, 2158.)

Koltroit, I. DL: Die ^Massannly-sc II. Berlin 1931.

Peters, J. P. and D. D. Va.v Slyke: Quantitative clinical Chemistry.

Vo1. II, Methods, London 1932.

PKirn, J. F.: Biochem. Z. 1929, 216, 219.

SciiOLAKPEK, P. F,; Skand. Arch. PhA'siol. 1938, 7S, 145.

ScHULF.K, K. and A. .Stasiak: Pliarmaz. Zentralhalle, 1928, 69, 513.
Ses'droy, j. Jr. and S. H. Liu: J. biol. Chem. 1930, 89, 133.
M'EN’.VESiARn, K.: Skand. Arch. Pliysiol. 1940. 83. 201.

VoK Weszei.sk Y, J.: Z. analyt. Chcin. 1900, 39, SI.

Wi.vKr.ER, L. W.: Z. angew. Chem. 1915, 28, 477, 494.

Winkler, L. W.: Z. anal}’!. Chem. 1934, 97, 18.

From the Gynecologic and Obstetric Clinic of Lund University, Sweden.
(Director: Professor AXEL WEST.MAX.)

A Parapharyngeal Method of Hn>ophysectoniy

in Rahhits.^

DORA JAGOBSOHN aod AXEL WESTMAN.

The observation of liypopln’sectomized animals is essential for
the study of the functions of the anterior lobe of the pituitary
body. For this reason special interest has been paid to the method
of extirpating the hypophysis. On account of the differences in
the anatomy of various animals it has been found necessary to try
out a particular technic for each species. Also in such cases where
the same route is chosen for exposing and removing the hj-pophysis,
the operation procedure has to be adapted to the animal species
concerned. For the rat in particular an excellent method has
resulted from the experiments of Smith (1930), Collip, Selye and
Thomp.sox (1933) and others. This method, in which the hypophy-
sis is removed by the parapharyngeal route, is comparatively
easy to learn. It is true that some have been successful in removing
rabbits’ hypophpes by buccal (White, 1933) or orbital (Firor,
1933) approach, but the fact that others (Newton, 1939, Cope
and Donaldsok, 1939) have found it necessary to use various
modifications shows clearly enough the difficulties of succesfully
applying these methods.

In an article of 1936 we gave a brief report of a parapharyngeal
method of operation for rabbits, agreeing in principle with the
method for rats.

The advantage of this method is that it enables a good survey
of the operation field. In addition, there is slight or no bleeding,
and no risk of infection from the pharynx as there always is with

' Received for publication 16 May 1940.

72 DOKA .TACOBSOIIK AND AXEL WE3TMAN.

tlie bticcnl mctliod. On histological examination of the contents
of tlie sella turcica and the hypophj'sis stalk attached no anterior
tissue could he foxind.

We have performed about 300 operations in conjunction with
different studies on the changes of the sexual organs (Westman
and jACOiisoiiN 193G, 1937) exhibited by rabbits after hypophys-
ectomy, and in all of them we have used the parapharyngeal
technic. Since we have been con-vinced of its many advantages,
we .«haU now give a more detailed description of the operative
procedure. The terms used are similar to those in Bensley’s
ftPractical Anatomy of the Babbit)).

The operation is performed on a table large enough to support
both clboAvs. The operation field is best illuminated by a head
lamp.

The animal is stretched out on its back and fastened. Its head
is fixed with a band hooked around the incisors of the upper
jaw and another ])and laid through the mouth across the upper
jaw (fig. 1). After cutting the hair, the anterior part of the neck is
washed witii alcohol. Tlie instruments and the hands of the opera-
tor arc washed, but it is not necessary to work under stcrilc
conditions,

Awfthriic. It is advisable to use an injectable anesthetic, at
least for a basal narcosis. AMien choosing the narcotic and dosage,
it .‘should be remembered that the narco.sis should not la.st more
than two to three hours. We find a combination of morphine
and urethane suitable. The animal is injected subcutaneously
with (1.00!) Gm. of morphine hydrochloride per kilogram of body
vreight, and half an hour later 0.5 Gm. of urethane (2 cc. of a 25
per cent solution) per kg. of body weight are injected intramuscu-
larly, distributed in two places. Pernocton (sodium, sec. butyl-/?-
bromallyl-barbiturat) has also been used in our operations, (0.3
cc. of a 10 per cent solution per kg, body weight intravcnou.sly.)
If ru'cessary ether can be administered during the course of the
operation.

Op-rnlivc trehnic. The skin and superficial fascine are divided
by .a mcdi.'m incision from the jaw to the sternum. A tracheotomy
b performed below the thyroid gland by a small tran.sverse inci-
sion, and an angular glas,'; tube is inserted into it (fig. 1). The inlra-
tr.'ir'ljea! part of the tube i*. about 2.5 mm in diameter. The left
subin.ixillary gland is lifted np and held to one side by a hook or
a iV an forcep-^. Tins gives clear e.xposnrc of the digastricus muscle.

HYPOPHYSECTOMY IN BABBITS.

the greater cornu of the hyoid, the hypoglossal nerve and the
stylohyoideus major muscle (fig. 1.). If palpation is done at the
place shown with the lower arrow in fig. 1, that is, at the origin
of the stylohyoideus major muscle, the pointed jugular process of

Fig. 1. The appearance of the operative field after incision of the skin and trache-
otomy.

the occipital bone can be felt. Knowing the features of the cranial
base (fig. 6), it is then easy to find the mastoid process and from
there the site of the basisphenoid. Two angular elevators, about
5 mm-, broad, are then inserted bluntly at the place indicated by
the upper arrow in fig. 1, which is situated lateral to the hypo-
glossal nerve and the greater cornu of the hyoid, medial to the
digastricus and caudal to the mylohyoideus. All vessels which

DORA JACOBSOHN AND AXEL WESTMAN.

Fig. 2. The field of operation after exposure of
the longus capitis muscles. A glass hook keeps
the esophagus and trachea to one side. The
point of the arrow lies against the median
vertebral vein which is compressed by the
elevator.

Fig. 3. The basisphenoid is exposed. The
point of the arrow lies against the cavernous
foramen.

appear during this manoeuvre are carefully pushed laterally and
dissection is continued in the medial and cranial direction until
the two longus capitis muscles and the interlying median vertebral
vein come into view. With the help of a hook to keep the esophagus
and trachea to one side and an elevator inserted at right angles to
the longitudinal axis of the longus capitis muscles, the two muscles
and the median vertebral vein are exposed (fig. 2). It is convenient
to use an angular glass hook, turned up at one end and connected
with a rubber tube which can be fastened to the operating table.
The elevator must be of such a size and shape that it lies well
against the basisphenoid, so that the median vertebral vein can

IIYPOPHTSECTOMY IN BABBITS. 75

be compressed against the bone with moderate pressure. For this
it is necessary that the elevator be held steadily, preferably with
the left hand keeping the elbow on the operating table. This makes
it possible to tear away the vein and detach the muscles from the
bone without any bleeding. The bone is exposed, as shown in fig. 3,

Fig. 4. Tho basisphenoid has been bored through and the hj'pophysia is exposed.

to such an extent that the entire basisphenoid and the transverse
spheno-occipital synchondrosis arc brought into distinct view.
The synchondrosis is generally easy to see, especially in small ani-
mals, because of its bluish colour, and it is always possible to find
it by palpation (fig. 5). If one follows the bony groove running
cranialward in the middle of the basisphenoid, one arrives at the
cavernous foramen (fig. 5), marked by the arrow in fig. 3, which
is relatively large in adult animals and which leads into the sphe-

DOHA JACOBSOHN AND AXED WESTMAN.

noid sinus. Since there is a copious flow of blood from this sinus,
it is stopped up before the basisphenoid is bored by means of small
bone wax plugs inserted carefully into the foramen. The bone
wax should be of such a consistency that the little rods, about one
centimeter long and one millimeter wide, are made sufficiently
pliable by the warmth of the operator’s hand, and yet it should

Fig. 5. The cranial base of the rabbit, seen from below.

be SO hard that it does not stick to the pincers used to insert it.
It is best to use a mixture of two-thirds paraffin with a melting
point of 55 C. and one-third cera alba. The rods can be niade
softer or harder as required by using paraffin with another melting
point. IVhen operating upon young animals, it is often impossible
to insert these wax plugs, nor is it necessary since the bleediug
generally stops of itself as the boring is continued and the spongiosa
removed.

UYPOPHYSECTOMY IN RABBIIB.

The tamponing ought to be finished by the time no more blood
oozes from the cavernous foramen. The basisphenoid is bored
through with a dental drill placed against the basisphenoid a few
millimeters cranial to the sphenooccipital synchondrosis and held
absolutely perpendicular. Fig. 6 shows the position and direction
of the drill in relation to the skull after the sella turcica has been
opened. The opened sella is also seen in fig. 4 where the hypophysis
is shown exposed.

Fig. 6. The position of the drill when the selln turcica is opened.

I^Tien the bone has been drilled through, the hypophysis is
seen clearly under the dura. The dura is slit wide open with a
cataract knife and the hypophysis sucked out under visual control
through a glass cannula connected with a water vacuum pump.
The bore hole is cautiously filled up with bone wax. The glass
cannula is removed from the trachea. It is checked that the respi-
ratory passages are not obstructed by blood or mucus (preferably
by means of a small suction tube sbghtly inserted into the trachea).
The wound is then closed with sldn sutures only.

The operation takes twenty to thirty minutes and can be per-
formed without any assistance. During dissection down to "the
skull base bleeding ought not to occur and the tamponing being
successful very little bleeding will take place during the whole
course of the operation.

If the animal does not begin to eat spontaneously the day after
the operation, it is good to inject about 30 cc. of a ten per cent
solution of glucose subcutaneously several times a day.

DORA JACOBSOHN AND AXEL WESTMAN.

Once the operator has mastered the technic, he will have hut
few deaths and, if one has healthy animals and tends them carefully
after the operation, the mortality can be brought down almost to
zero.

Summary.

A technic of hypophysectomy in rabbits is described. The
parapharyngeal route is taken and the operation done according
to the principles of hypophysectomy in rats, but modified to suit
the anatomy of the rabbit. The method has great advantages. It
enables direct inspection of the field of operation. The operative
risk is slight. The mortality is extremely low.

This work has been aided by a grant from wStiftelsen Therese
och Johan Anderssons Minneft, Stockholm.

References.

CoLLiP, I. B., H. Serve, and D. L. Thompson: Virchows Arch. 1933.
290. 24.

Cope, 0., and 6. A. Donaldson: Endocrinology 1939. 24. 475.
Fieor, W. M.: Amer. J. Physiol. 1933. 104. 204.

Newton, W. H.: Endocrinology 1939. 24. 468.

Smith, P. E,: Amer. J. Anat, 1930. 45. 205.

Westman, a., and D. Jacobsohn: Acta obstet. gynec, scand. 1936.
16. 483.

— , Ibidem 1937. 17. 13.

White, 117, E.: Proc, Roy Soc. London B. 1933, 114. 64.

From the Biochem. Dep. of Karolinska Inst., Stockholm.

SuU'apyridiiie in Secretin Stinmlated Pancreatic
Jnice and Bile of Cats.^

A. TAYLORS nnd G. AGREN.

For some time past, part of the interest of this laboratory has
been devoted to work with secretin and its physiological and thera-
peutical values (Agren 1939; Hammarsten 1939). In this article
we present the results of experiments combining the use of se-
cretin and sulfap}Tidine in connection with the presence of the
latter in the pancreatic juice and bile. It would seem that the
pancreas and liver in excreting this drug into the duodenum with
the pancreatic juice and bile display the same type of activity
as has been noted with the excretion of injected urea and uric
acid (Agren 1935).

Amounts of sulfapyridine or sulfanilamide have already been
found in pancreatic juice and bile. Marshall c. s. (1937 a)
commented shortly on the presence of the drug in pancreatic
juice and saliva of the dog, Carryer and Ivy (1939) have found
it in the secretions of all digestive glands of the dog, and
Bettman and Spier (1939) observed it in human bile. Om:
work confirms these results and approaches the problem from
a different angle — a combination treatment of secretin and
sulfapyridine.

We have also interested ourselves in the change of the drug
from free to bound form within the system. In this connection
Stewart (1939), Kourke and Allen in working with rabbits

* Received for publication 27 May, 1940.

® Exchange Fellow, Scandinavian-American Foundation, 1939 — 40.

A. TAYLOR AND G. IgRBN.

conducted in vivo experiments to show that the acetylation of
sulfanilamide takes place in the liver, and Klein and Harris
(1938) previously found the same thing in in vitro experiments
with rabbit liver slices.

All work has been performed on cats, in whom the pancreatic
juice and bile were collected separately to determine the drug
content of each. One impelling reason for using cats rather than
other animals was that they are similar to man in the change of
the drug inside the body from free to bound form (jMarshall
c. s. 1937 a).

Methods. The method used in determining the amount of
sulfapyridine in blood, pancreatic juice and bile is essentially
that described by IMarshall c. s. (1937 b and 1938). The readings
were made in a Pulfrich Photometer with an S50 Jena filter and
compared with standard curves. For total drug analysis in all
fluids, alkaline hydrolysis was used being slightly more effective
than acid after the method of Baines and Wien (1939).

All injections of sulfapyridine and secretin as well as the re-
moval of blood were made in the femoral veins.

Since the pancreatic duct in the cat does not allow of cannula-
tion, pancreatic juice was collected on tightly-rolled filter-paper
wads immediately on its secretion into the duodenum (Agren
1935). These were boiled out with water four times to a total
volume of approximately 75 cc, and filtered. The solution was
then evaporated in vacuo to about 5 cc, or the approximate volume
of the pancreatic juice collected. This was washed out and diluted
to 10 cc. Then to determine the volume of pancreatic juice, a
one cc portion of this solution was made acid with 10 cc N/lOO
HCl and titrated against N/lOO NaOH. Since the bicarbonate
concentration of evenly flowing pancreatic juice may be taken
as K/10, the total volume can be calculated.

Bile was collected by cannulation of the Ductus choledochus
after previous ligation of the Ductus cysticus to ensure the flow
of hepatic bile only. To precipitate the bUe pigments, which inter-
fere with color readings, the following method was used. Approxi-
mately 0.3 cc bile were diluted to 3 cc with distilled water and made
slightly alkaline with 0.5 cc N/10 NaOH. Then 0.5 cc of a 4.5 %
solution of ZnSO^ were added to precipitate the proteins and with
them some of the closely associated pigments. 0.6 cc N/1 NaCOs
were then added, immediately followed by 0.5 cc N/1 BaCl-.
This fresh precipitate of BaCOa brings down the remaining bile

FULFAPYIUDIKE IK SECKETIK STI.MULATED PANCREATIC JDICE. Si

pigments, and wlicn filtered after 10 minutes a clear, colorless
solution is obtained. There is no ad.sorbtion of the drug during
this procedure.

The cats, jinder urethane narcosis, were given suIfapjTidiuc
intravenou.«lv (about. 100 mg per kilogram bod}' weight) in 4 por-
tions at o—lO minuto.s intervals. Individual cat differences were
such that the minimum time, for these injections varied, but less
than 5 minutes tended to prodnee vomiting. The sulfa p}Tidine
injected in the first cats was in a .specially prepared form (15 — 25 %
solutions) intended for intravenous injection. It was obtained
from Pharmacia. Stockliolm, under the name Septiglucon. Un-
fortunately, analysis showed that only 5 % of the sulfap}T:idine
in solution was in the free form, a concentration too low to permit
observation of change in form from free to bound. However,
a 20 mg. pr-r cc solution of sulfa pjTidino, 45 % of which was un-
bound, wa.s obtained by dissohdng 1 gram of the rocrystallizcd,
powdered drug in 50 cc of hot 12 % glucose solution. This was
more than adequate to determine the difference between pancreas
and liver actinty in changing the form of the drug.

In all cats, the first injection of secretin (Pancreotest, from
.A.stra, Sodertaljc, Sweden) was made between 10 and 20 minutes
after the In.st sulfapyridino injection. The amounts of secretin
injected (approximately 0.3 mg per kulogram body weight) were
calculated to produce mn.ximnl secretion, and continued doses
during the coiursc of the experiment gave a continuously even
flow juice. Immediately before the first secretin treatment, 1.5 cc
of blood were taken from the femoral vein, so that the standard
could be had ngain.st which the pancreatic juice and bile could
be compared. Another blood sample was taken at the end of the
experiment, and in some cases one or more were also taken during
the course of the experiment.

In Table 1 may be .seen the general outlookof the experiments.
In ail case.s, “experiment’' refers to the period over which secretin
was injected. The time between the Inst sulfnp}T:idine injection
and the first one of secretin Ls not the same in all cases, although,
the variance is no greater than ten minutes.

Tabic 2 compares rate of flow of pancreatic juice and bile with
the concentration of the drug. In Cats HI and IV, the juices were

fractionated (represented by a, b, c, d) III in half hour

period.s, and IV in fifteen minute periods. The values were cal-
culated for periods of 30 minutes.

6 — i0132S. Acta phya, Scandinav. VoUl.

A. TAYLOR AND Q. AqREN.

Table 1.

Concentrations of Sulfapyridine in Blood, Pancreatic Juice and Bile

after Secretin.

Cat I

Amt. drug given per
kg body wt . . .

80 mg

96.5 mg

112 mg

110 mg

92 mg

81.6 mg

Blood cone, before
exp

13.8 mg^

9.66 mg?S

10.6 mg5{

15.8 mgjs

12.0 mg%

Blood cone, after
exp

5.8 >

3.8 >

2.7 mgj^

2.6 >

7.8 >

8.8 »

Conc’n in Pan. Jnice

5.11 >

5.6 >

5.0 >

1.9 >

7.6 »

12.6 >

Conc’n in Bile . . .

19.2 >

359.8 >

102.5 >

34.1 »

59.6 >

Length of Experiment

60 min.

90 min.

60 min.

45 min.

Table 2.

Concentrations of Sulfapyridine in Relation to Rate of Flow of
Pancreatic Juice and Bile.

Pancreatic Jnice

Blood
Conc’n of
Snlfap.
mg/100 cc

BUe

Rate of
Flow

cc/30 min.

Cone, of
Snlfap.
mg/100 cc

Rate of
Flow
cc/30 min.

Cone, of
Snlfap.
mg/100 cc

Cat I

2.2

5.11

5.8

Cat n

2.6

5.6

5.8

a) S.6

5.6

6.8

2.4

510

Cat ni . . . . <

b) 1.7

5.2

1.9

328

c) 2.4

4.4

2.7

1.6

244

a) 4.0

1.66

—

1.68

136

Cat rv .....

b) 4.44

1.69

2.6

1.82

121

c) 3.06

2.82

2.6

1.10

69.8

0.7

1.66

19.2

Cat V

4.0

7.6

7.8

0.89

34.1

Cat VI

3.6

12.6

12.1

1.0

59.6

The figures for the change in form of sulfapyridine in pancreatic
juice and bile are given in Table 3.

The extremely high concentration of sulfapyridine in the bile
of Cats III and IV coincides with the use of a slightly different
drug solution.

SDIiFAPTRIDINE IN SECRETIN STIMULATED PANCREATIC JUICE. 83

Table 8.

Change of Sidfapgridine from Free io Bound Form in P. J. and Bile.

Ratio of free
to bound drug
injected

Ratio of free
to bound drug
in Pan. Juice

Ratio of free to bound
drug in Bile

Cat IV

13/87

..... ..

0.41/99.6 (after 60 min.)

Cat Y

5/95

6.6/93.7

4/96 (after 45 min.)

Cat VI

45/55

46/54

31/69 (after 45 minutes)

Results.

The blood level ■which “was used as the standard of comparison
varied in every cat, but exhibited a rapid decrease in concentra-
tion at first, followed by a very gradual decrease over a longer
period. The concentration of sulfapyridine in pancreatic juice
was slightly lower than in blood in every case except Cat '\C[,
where this high concentration is probably due to an error in meas-
uring the volume of juice. The result agrees with pre'vdous find-
ings by other experimenters -who "were using secretions of pan-
creatic juice in dogs.

The concentration of the total drug in the bile was extremely
high in every cat, much higher than would be expected from
the work of Carryer and Ivy (1939) on dogs.

An increased flow of digestive juices can definitely be stated
not to decrease the concentrations of sulfapyridine. At the same
time, it is not possible to determine from our results if the con-
verse of this is true. However, it would seem fairly clear that the
concentration of drug in both pancreas and liver secretions is
relatively independent of the rate of flow of the juices, and more
dependent on the concentration level in the blood. Disregarding
the concentration per unit volume of the drug in the pancreatic
juice and bile, it is ob-vdous that a very large total amount of the
drug can be washed out on secretin stimulation of the two organs.

In connection •with the change of sulfapyridine from the free to
the bound form quantitative experiments were attempted in
Cats IV and V -with only qualitative results due to the insufficiency
of free sulfapyridine in the injected preparation. The results
nevertheless indicated that the relative percentage of free drug
excreted in pancreatic juice remained unchanged while there
was little or none excreted in the bile. In Cat VI, the 45 % free

A. TAYLOE AND G. IgKBN.

drug solution was injected and 15 minutes after the last injection,
the bile and pancreatic juice were collected over a half hour
period. Analysis showed that in the pancreatic juice 46 % of
the total drug was still free, or that there had been no change.
In the bile however, only 31 % of the total drug was unbound,
indicating the acetylation of 30 % of the free drug over a period
of 45 minutes. It is very likely that acetylation of the drug
takes a longer time due to the rapid washing when larger flows
of bile are stimulated by secretin injection.

Summary.

The combined treatment of cats with sulfapyridine and secretin
is described. In conformance with other investigations, in which
secretin was not used, the concentration of sulfapyridine in
pancreatic juice was found to be just lower than that of blood,
but an extremely high concentration of the drug was found in
the bile. The drug is not diluted by the increased flow of juice
either from pancreas or liver and tends to be independent of rate
of flow. Blood concentration of sulfapyridine is more important.
While there was no change in the percentage of free drug excreted
in the pancreatic juice, 30 % of the free drug had been bound
when excreted in the bile, indicating acetylation by the liver.

References.

Baines, E. J. and R. Wien, Quart. J. Pharm. 1939, 12 , 4.
Bettsian, R. and E. Spiee, Proc. Soc. Exp. Biol. N. Y. 1939, 41 , 463.
Caeryee, H. M. and A. C. Ivy, J. Pharmacol. 1939, 66 , 302.
HAAIMA.RSTEN, E., J. of Mt. Sinai Hosp. 1939, 6 , No. 2.

KiEiN, J. R. and J. S. Harris, J. Biol. Chem. 1938, 124 , 613.
Maesecall, E. K., K. Emerson and W. C. Cutting, J. Pharmacol.
1937 a, 61 , 196.

— , J. Biol. Chem. 1937 b, 122 , 263.

Marshall, E. K. and C. Liohtfeld, Science. 1938, 88 , 85.

Stewart, J. D., G. M. Roueke and J. G. Allen, Surgery, 1939, 5 , 232.
Ageen, G., Biochem. Z. 1935, 281 , 358.

— , J. Physiol. 1939, 94 , 553.

From the Pharmacological Department, XJniversity of Lund, Sweden.

Some Qiiaiititatiyc Data on the Antagonism
between Piperido-Mctliyl-Benzo-Dioxane
(933F) and Adrenaline."

N.-O. ABDON and S. O. HAMMARSKJOLD.

According to a modern aspect on the mode of action of ad-
renaline, the site of action of this drug is attributed to certain
receptors. The action of the drug involves two phases, the com-
bining of the drug molecules with the receptors and the exertion
of action after they have combined. This hypothesis, developed
by Ci/ABK, is supported by the fact that the concentration-action
curve of adrenaline follows Langmuir-Hitchcock’s law- (vide
Clark, 1936). Drugs wliich are specifically antagonistic to ad-
renaline are considered to have the ability of combining with
the adrenaline receptors, thus blocHng receptors to this drug,
but unlike adrenaline the antagonistic drugs have no or little
power of gi%dng an effect after being combined with the recep-
tors. In the case of certain antagonists it has already been made
probable that the affinity between the receptors and adrenaline,
respectively between the receptors and antagonist, is governed
by the law of mass action, t. e. the presence of the antagonist
does not alter the .shape of the concentration-action curve, which
still follows Laxghgir-Hitchcock’s law.

^ Received 6 June, 1940.

s k - I = y — ; k is a constant, x is tke concentration of tlie drug prodncing

100 —y

the effect y; y is calculated as percentages of the maximal effect which can be
produced by the dmg in qnestion.

N.-O. ABDON AND S. 0. HAMMAKSKJSlD.

The experiments of Bacq and Febidericq (1935) suggest that
Clark’s hypothesis on drug antagonism may he applied to the
antagonism between 933 F and adrenaline. Their results seem
to indicate that 933 F inhibits the action of adrenaline by inter-
fering with the specific receptors. The authors have, however,
studied the antagonism only within a narrow range of concen-
trations of the drugs, which must be said to limit the possibilities
of drawing conclusions.

In his article in Physiological Reviews (1937), Rosenbltjbth
has another aspect on this antagonism. He tries to explain the
fact that 933 F antagonizes adrenaline added to an organ, while
it does not inhibit the effect of the stimulation of the sympa-
thetic nerves of the same organ. He assumes that 933 F renders
the cells impermeable to adrenaline, thus blocking its passage
to its site of action, while the transmitter of sympathetic nerves
is liberated within the cells in close relation to the site of action.
According to this view, 933 F does not exert any action on the
adrenaline receptors themselves. A confirmation of Bacq and
Fr^dericq’s opinion may thus have a certain interest, and we
have, therefore, studied the quantitative relations of the antag-
onism of 933 F to adrenaline in experiments on perfused ears
of rabbits.

In the case of the antagonistic drugs which are supposed to
act on the specific receptors the quantitative data follow a simple
expression, developed by Gaddum (1937) on the basis of Lang-
muir-Hitchcock’s law and originally applied to the antagonism
between acetylcholine and atropine:

Ki [Adr] = (1 -f Kj [933 F]“) ^ — ;

Ki and Kj are constants; y is the effect which is produced by
the concentration of adrenaline [Adr] in the presence of the
concentration of [933 F]. The exponent, n, expresses the
number of molecules of the drug which combine with one
reeeptor. In our case the value of n was found to be 1,
i. e. one molecule of the drug combines with one receptor.
As we always determined the amotmts of adrenahne giving

SOME QUANTITATIVE DATA.

the same effect, the factor y was kept constant. The expres-
sion can then be simplified to:

[Adr] = (1 + Ks [933 F]) • a;

The experiments were made on rabbits’ ears, which were per-
fused through the artery with Ringer’s solution. By means of
Mariotto’s flasks the perfusion pressure was kept constant at
350 mm. of water. The arrangements made it possible to shift
from Ringer to 933 F solutions without interrupting the per-
fusion and without changes in the pressure. The rate of per-
fusion was measured with Rothlin’s “Ordinatenschreiber”,
which automatically records the number of drops per minute.
The various amounts of adrenaline, always dissolved in the same
amount of Ringer solution, wore injected in the connecting tube
immediately before the cannula, inserted into the artery. In
this arrangement those amounts of adrenaline were determined
which produced a certain decrease of the drop rate before and
after the addition of various amounts of 933 F. The results are
seen from the table below.

The values experimentally found arc in good accordance with
values, calculated according to the expression above. The ex-
periments, therefore, seem to indicate that the molecules of
adrenaline and of 933 F react with the receptors in accordance
with Gaddd.m’s expression. Thus, the experiments support the
observations of Bacq and Fredericq.

Concentration of
933F per ml of
perfusing fluid)

0.01

0.1

100

1,000

Experiment 1

/ of adrenaline pro-
dneing the same
decrease of drop
rate

a) observed . .

0.003

O.OOG

0.012

O.l

1.0

b) calculated. .

0.012

0.09

0.9

Experiment 2

a) observed . .

3.0

300

2,500

•

b) calculated . .

•

3.6

250

2,500

•

N.-O. ABDON AND S. 0. HAMMARSKJOLD.

Eeferenees.

Bacq, Z. M. and H. Fredebicq, Aicli. int. Physiol. 1935, 4^0, 454.
Clark, A. J., Heffters Handbuch exp. Pharm. 1936, Erg. bd. IV,
pp. 185 foil.

Gaddxjm, J. H., J. Physiol. 1937, 89, 7 P.

EosENBLtTETH, A., Physiol. Rev. 1937, Tt: 4, 522.

Aub der pharmakologischen Abteilung des Karolinischen Instituts

zu Stockholm.

Wirkung der Sauerstoffatmung anf die Atmungs-
steigerung bei Carotisabklemmung.^

Von

TORE RUDBERG.

(Mit 1 Abbildnng im Text.)

In einer friilieren Arbeit (Rudberg 1938) wurde gezeigt, dass
die ventilationssteigernde Wixkung der Carotidenabklemmung
bei Katzen und Kanincben wesentUcb verstarkt wird, wenn der
arterielle Blutdruck des Tieres unter einen gewissen kritischen
Wert, der individuell verscbieden ist, gefallen ist. Es vrurde ferner
die Vermutung ausgesprocben, dass dieser atmungserregende
Effekt von einer Hypoxamiereizung des Sinus caroticus abbangen
konnte. Es -war desbalb von Interesse zu priifen, ob Einatmung
von Sauerstoff den erwahnten Effekt beeinflussen wiirde.

Solcbe Versucbe sind jetzt an vier Katzen ausgefiibrt worden.
Die Tiere mirden mit Uretban narkotisiert, worauf Blutdruck
und Atmung registriert vmrden. Die Atmung v*urde in einem
Ealle mit Pneumograpb, in den iibrigen dagegen quantitativ mit
Korperpletbysmograph (vgl. v. Euler und Liljestrand 1936)
aufgezeicbnet. In zvrei von diesen Fallen ist die Carotisabklem-
mung durcb spezielle Klemmen, die die ganze Zeit rings um die
Gefasse in fixierter Lage befestigt ■waren, durcbgefubrt vorden.
Hierdurcb vuirde die Ziebung bei der Abklemmung vermieden,
und der Eingriff konnte geschehen, obne dass der Pletbysmograpb
geoffnet werden musste. Die Tiere atmeten durcb Miillerventile.
Durcb Blutentnabme 'wnrde der Blutdruck gesenkt, bis die
Carotidenabklemmung kraftige Vermebrung der Atmung be-
vrirkte. Jetzt wurde statt Luft Sauerstoff inspiriert; nacb einer

* Der Redaktion am 22 . August zugegangen.

TURE EUDBERG.

iRlespiratioia

L uft

Sauerstoff

BlutdrucK

mm Hg

^0-

j' X I Af ao-

Signal 0“

fo’ Zeit

Katze 1. Die Kurven bedeuten von oben VentUation, Blutdrnck, Signal und Zeit
(10 SekO- Durch Carotisabklemmung (zwischen 4 nnd X) vrird die Ventilation
bei Lnftatmung nm 48 Droz., bei Saaerstoffatmung um 11 Proz. vermehrt.

Vorperiode von einigen Mnuten vrurden die Carotiden noclimals
abgeldeimnt und zwar walirend etwa derselben Zeitspanue "wie
in dem friiberen Versucb (gewobnbcb 1 — 2 Llinuten). Hierbei
ist der Blutdruck so gut wie unverandert geblieben. Der Versucb
ist mebrmals an demselben Tier ausgefubrt -worden.

Es bat sicb in samtlicben Fallen gezeigt, dass die Atraungs-
steigerung nacb CarotidenabMemmung wesentlicb langsamer
eintrat und kleinere Werte erreicbte, vrenn das Tier Sauerstoff
geatmet batte als bei Luftatmung. Der Unterscbied ist in ver-
scbiedenen Fallen ungleicb gross, aber immer ganz deutlicb ge-
■wesen. Die Abb. 1 gibt einen typiscben Versucb wieder, und in
der Tabelle 1 werden die quantitativ ermittelten Werte zusammen-
gestellt.

Es scbeint also, als ob die Sauerstoffatmung durcb bessere
Sattigung des Hamoglobins mit Sauerstoff die atmungssteigernde

WIRKUNG DER SADEBSTOPFATMUNG. 91

Tab Clio 1.

Vcrsnchstier

Inspirations-

laft

Blut-

druck

lum

Ventila-
tions-
Ver-
mebrang
in Proz.

Zoit, irfihreiid
■welclier der
'Vcrgloich
gescbchen ist

Bemerkungen

Katze 1 . .

j Zimmcrluft

1 Saucrstoff

I Zimmerlnft

1 Saucrstoff

• 40 Sek.

Carotidenabklem-
mung mit spe-
zicllen Klemmen

Katze 2 . .

1 Zimmcrluft
\ Saucrstoff

} 20 Sek.

> > >

Katze 3 . . ;

f Zimmcrluft
\ Saucrstoff

J- 60 Sek.

Wirlaing der Carotidcnabkleintiiimg entgegen-wirken kann. Dass
das Blut bei den nicdrigen Blntdruckwerten, die bier in Frage
kommcn, nicht mit Saucrstoff gesiittigt ist, \mrde aucb direkt;
festgestellt. Es wurden zu diosem Zwecke arterielle Blutproben
teils bei cinem Blutdruck, wo die Carotidenabkleinmung nur wenig
die Ventilation stcigerte, teils aucb unterbalb des kritiscben Dxuk-
kes entnoniinon und nacli van Slykcs manometrischer Methode
analjsierfc. Die Tabelle 2 gibfc die Ergebnissc,

Tabello 2.

Tersuchsticr

Blut-

dfnek

Saucr-

stoff-

gehalt

Yol.

Sancr-

etoff

sSttigung

'SVirkung der
Carotiden-
abklemmung
auf die

Bemcrknng

•*

in %

Atmnng

16.8

100

schwach

16.8

100

sckn’acli

kritiseker Druck

Katze 4 . ,

■svakrscheinlich

16.0

gut

zrriseken 68 und

56 mm

15.5

93.4

gat

Katze 5 . .

f 60

1 41

18.1

16.0

96.4

82.8

sekwack

gut

Aus den Versucben muss gescblossen werden, dass wenigstens
ein wesentlicher Teil der Ventilationsvermehiung, die durcli Caro-
tidenabklemmung bei der Katze mit relativ niedrigem Blutdruck
erbalten wird, durch eine Hypoxamiereizung zustande kommt.
Das Ergebnis ist in guter Dbereinstimmung mit dem Befund,

TURK BDDBBRG.

dass bei der Carotisabklemmung eine vermebrte Anzahl von
Aktionspotentialen aus den Nervenfasern der Chemoreceptoren
des Sinus caroticus nacbgewiesen werden kann (v. Euler, Lilje-
STRAND und ZOTTERMAN 1939).

Zusammenfassung:.

An der Katze Tvird gezeigt, dass der ventilationsvermebrenden
Wirkung der Carotidenabklemxnung bei relativ niedrigem Blut-
druck durcb Einatmung von Sauerstoff entgegengewirkfc werden
kann. Ein wesentlicber Teil wenigstens der Atmungssteigerung
muss desbalb durcb H 5 rpoxamiereizung des Sinus caroticus zu-
stande kommen.

Literatur.

V. Euler, TJ. S., and G. Liljesteand, Skand. Arcb. Pbysiol. 1936.
74. 101.

V. Euler, U. S., G. Liljestrand and Y. Zotterjun, Ebenda. 1939.
8S. 132.

Rudberg, T., Ebenda. 1938. 79. 8.

From the Phnrmacological and Physiological Departments
of the Karolinska Institutct, Stockholm.

The Effect of Carotid Siinis Deneryatioii
on Respiration during Rest."

U. S. V. EULER and G. ULJESTRAND.

The discovery by Hfa'mans, Bouckakrt and Dautrebande
(1930) that, the chcinorcccptors of the carotid body are stim-
nlated by h 37 )oxemia, as well as b_v h}'])crcapnia, thereby in-
creasing respiration reflcxly, raised the problem whether such
reflexes play any rule under physiological conditions or may be
considered onl}’ as part of an emergency mechanism. It was
soon generally agreed that oxygen lack stimulates respiration
practically only over the .sinus mechanism, whereas opinions
still differ with regard to carbon dioxide. Thus v. Euler and
Liuestraki) (193G) and IIeymans and Bouckaert (1939) came
to the conclusion that carbon dioxide influences respiration nor-
mally bj" the reflex mechanism mentioned; Schmiut and Comroe
(1938), on the other hand, consider that the sinus mechanism
hardly tnkes any part in this regulation. Investigations on the
action potentials from the sinus nerve have shown conclusively,
as far as we can judge, that even a small increase in the desatura-
tion of the hemoglobin or of the carbon dioxide tension of the
blood will cause a considerable rise in the number of potentials
(Samaan and Stella 1935, v. Euler, Liljestrand and Zot-
TERMAE 1939). It was also observed that such potentials are
elicited already at a hemoglobin saturation with oxygen of about
95 per cent or a carbon dioxide tension of about 30 mm respec-
tively. These observations would seem to indicate that th.e re-
flexes concerned probably act under ordinary conditions, but

* Received 3 Septorabor 1940.

U. S. V. EULER AND G. LILJESTRAND.

they tell us nothing with regard to their quantitative signif-
icance.

The simplest way of determining this is undoubtedly to com-
pare respiration before and after denervation of the sinuses. If
respiration during rest is normally to some extent stimulated
from the sinus region, then it is to be expected that denervation
will reduce the ventilation. The possible compensation from
the aorta nerves, analogous to that which occurs with regard to
the blood pressure effect of sinus denervation, seems to be of
little significance for the respiration. There is, however, another
complication, since the denervation has an indirect stimulating
effect on respiration, which is quite independent of any normal
direct chemoreflex over the respiratory centre. It was found
by V. Euler and Liljestrand (1935) that clamping the caro-
tids in the dog resulted in an increased metabolism. The rise
in the oxygen consumption varied considerably; on an average
it was 11 per cent. Mertens and Eein (1938) confirmed this
result, though the increase observed by them was smaller. Hahn
(1940), on the other hand, after section of both the sinus and aorta
nerves in the dog, found a rise of 10 — 35 per cent in the oxygen
consumption. As is well known, increased metabolism causes a
rise in ventilation, and there is no reason why this should not be
true in cases where the rise in metabolism is brought about by
sinus denervation. It is to be expected, however, that the al-
veolar carbon dioxide tension will not rise appreciably with a
moderate increase in metabolism. If, now, denervation of the
sinuses does away with reflexes which normally stimulate respi-
ration, then the increase in ventilation from the rise in metab-
olism may become more or less compensated, or the result may
even be increased ventilation. A definite index of the signif-
icance of the reflexes concerned will, however, be a rise in the
alveolar carbon dioxide, indicating that the centre alone is im-
able to maintain it at the usual level. This increase above the
value before denervation will therefore give a clear expression
of the role of the chemoreflexes under physiological conditions.

A decrease in ventilation after denervation of the sinuses has
been observed by several investigators. Thus, in 4 cats under
chloralose anesthesia, Selladurai and Weight (1932) found ah
average decrease of 22 per cent. In 6 decerebrated cats the cor-
responding average lowering was 33 per cent. It must be ob-
served, however, that the carotids were open only in one of these

EFPECT OF CAEOTID SINUS DENERVATION. 95

cases (witli a decrease in ventilation of only 6 per cent). Since
the ligaturing of the carotids gives rise to oxygen -want, vdth a
corresponding increase in ventilation, which is abolished by the
denervation, these last-mentioned experiments give no satisfac-
tory evidence as regards the normal participation of the sinus
mechanism in the regulation of respiration. In dogs, cats and
rabbits Witt, ELatz and Eohn (1934:) observed a respiratory
standstill, and death after careful denervation of the sinus re-
gion. They conclude that a tonic influence on respiration is
exercised from the sinus region, v. Euler and Liuestrand
(1936) found in anesthetized cats (chloralose) a rise in the alve-
olar carbon dioxide of some 1.5 per cent, indicating a decrease in
ventilation of about 25 per cent. In anesthetized dogs Gesell and
Lapides (1938), as well as Schmidt, Comroe and Dripps (1939),
observed depressed respiration in consequence of the denerva-
tion. Schmidt, Dumke and Dripps (1939), experimenting on
lightly anesthetized dogs, whose carotid pressoreceptors had been
divided, while the chemoreceptors were left functioning, found
that blocking of the sinus nerves with prokain during oxygen
inhalation did not consistently affect respiration. It ought to
be remarked that the dogs had received morphia, which is known
greatly to affect the respiratory centre.

The evidence quoted above has recently been reviewed by
Schmidt and Comroe (1940). They maintain that anesthesia
may greatly influence the respiratory centre, whereas the chemo-
receptors remain effective; they further point out that the
denervation may be accompanied by trauma, affecting respira-
tion, and that the rise in blood pressure following the denerva-
tion may be responsible for changes in breathing. There is also
the possibility that a decreased ventilation after sinus denerva-
tion may, wholly or in part, be the consequence of some oxygen
want which will stimulate respiration over the sinus mechanism,
but not directly on the centre. According to Schmidt and Com-
roe the above-mentioned experiments by Schmidt, Dumke and
Dripps are free from these objections.

It is quite clear that the use of an anesthetic will depress the
respiratory centre more or less. We think, however, that the
effect can be greatly reduced if a suitable anesthetic is chosen
and the dose given is not too large. It is probably possible to
find a point where the cortex is eliminated, though the centres
of the medulla are hardly affected. The animal will not then

U. S. V. EULER AND Q. LIUESTRAND.

display the great lability of the untrained 'unanesthetized or of
the lightly anesthetized animal, and small sensitive stimuli ■wiU
have no influence on respiration. If the depression of the respi-
ratory centre is noticeable, one must expect that the alveolar
carbon dioxide tension will rise. We have shown before (1936)
for cats — and similar results were obtained for dogs, though
they were not reported in detail — that the alveolar carbon
dioxide was held for hours about 6 — 6 per cent of an atmos-
phere, when the animal had been anesthetized with chloralose
(6 — 10 eg per kg of body weight). This completely tallies with
the values found by Eoos and Romijn (1937) for the trained
unanesthetized dog. As for cats no direct observations without
anesthesia are known, but there is good reason to believe that
the alveolar carbon dioxide tension is similar to that of the nor-
mal dog. If sleep is beginning to become superficial, a drop in
the carbon dioxide tension marks that the experimental condi-
tions act as a sensitive stimulus. A light anesthesia, as used by
Schmidt, Dumke and Drifts, may therefore easily involve spe-
cial difficulties in obtaining constant results.

There can be no doubt that a trauma may greatly influence
respiration. The type of trauma involved in our experiments
will always — with the exception of transitory effects of
very short duration — produce a stimulation of respiration.
Consequently a depressing effect of the sinus denervation may
to some extent become obscured by the trauma. That the de-
crease in respiration after the denervation which has been ob-
served by several investigators is to be explained by the effect
of the unspecific trauma is in our opinion very unlikely, and no
proof of this assumption has been given.

With regard to the possible effect on respiration of the rise in
blood pressure following the denervation, the evidence available
does not indicate any such effect. Heymans and Bouckaert
(1930), after having cut the depressor and sinus nerves in'dogs,
observed no change in respiration after the intravenous injection
of 0.3 mg adrenahne, which produced a considerable rise in the
blood pressure.

The rise in alveolar carbon dioxide after sinus denervation was
at first interpreted as a proof that carbon dioxide stimulated
respiration, in part reflexly, under normal conditions (Euler
and Liljestrand 1936), The possibility that the effect might
to some extent be due to the absence of a stimulation from oxy-

EFFECT OP CAROTID SINDS DENERVATION.

gen lack was excluded by the observation that a corresponding
rise in alveolar carbon dioxide was observed, even if the animal
was breathing oxygen. The results quoted above concerning the
action potentials in the sinus nerve have demonstrated, how-
ever, that even a very small degree of oxygen lack may stimulate
the nerve endings. It is therefore quite likely that a simultaneous
effect of oxygen lack and carbon dioxide on the sinus mechanism
occurred in our earlier experiments, though carbon dioxide alone
was responsible for the result in the controls mentioned. In order
to determine the effect of carbon dioxide alone it became desir-
able to make experiments while the animal was breathing oxy-
gen.

Methods.

We have used cats and dogs in our experiments. They were
usually anesthetized with chloralose (0.06 — 0.10 g per kg of body
weight). In the dogs a small amount of barbiturates (0.1 ml
somnifen per kg) was added, in order to reduce shivering. Some
experiments were performed with cats decerebrated under ether
anesthesia. Since it was necessary to maintain the carotids in-
tact, they had to be kept open, and this involved a great
risk of hemorrhage. Among several methods tried, we found
the intravenous injection of sangostop (2 ml of a 6 per cent so-
lution per kg of body weight) before the decerebration the most
effective means of preventing bleeding. The animal breathed
through Muller or Lov4n valves. Alveolar air samples were ob-
tained through a fine catheter, inserted in the side-tube of a
tracheal cannula. The lower end of the catheter reached the
bifurcation of the trachea, and an airtight connection was estab-
lished with the cannula. A few ml of alveolar air were drawn
into a sampling tube at the end of each respiration. Ventilation
was recorded quantitatively with the aid of the body plethys-
mograph described in an earlier paper (Euder and Liljestband
1936). Denervation of the sinuses was made by mass ligation
of the carotid body, the bulk of tissue between the external and
the internal carotids being ligatured, after the region concerned
had been carefully dissected free before the experiment started.
The reaction of the chemoreceptors before and after denervation
was tested by letting the animal inspire a gas mixture of 7 per
cent of oxygen in nitrogen.

7 — ^01323. Acfa pliys. Scandinav. Vol.1.

U. S. V. EULER AND G. LTLJESTRAND.

Results.

The following abbreviated record will serve as a type experi-
ment.

6. 3. 1940. Cat, 4.8 kg, receives under ether anesthesia 35 ml
of a 1 per cent solution of chloralose intravenously. Tracheal
cannula. Sinus region dissected. Animal in body plethysmo-
graph. Blood pressure recorded from femoral artery.

Time

Ventilation

Eespiration

Alveolar carbon

Blood pressure

1. per min.

rate per min.

dioxide per cent

mm Hg

10.47

0.G8

13.9

6.40

141

10.5G

10.57

O.CO

Inspires oxj’gen.

14.0

1.52

11.03

0.G5

13.2

C.GO

1.54

11.11

11.13

0.70

Inspires air.

0.7 G

14.0

14.7

6.57

166

176

11.27

0.7G

Inspires oxygen.

15.0

6.40

180

11.31

0.C8

14.2

173

11.34

11.3G

0.73

Inspires air.

14.G

6.40

172

11.37

0.74

14.7

180

11.39

0.78

15.4

181

11.45

7 per cent of oxygen in nitrogen gives the typical increase in ventilation,
Denervation of both sinuses.

Inspires air.

12.18

0.55

12.3

7.53

223

12.22

7 per cent of oxygen in nitrogen gives practically no
lation.

Inspires air.

increase in venti-

12.38

12.45

0.57

12.C

7.82

170

13.10

13.14

0.59

Inspires oxygen.

12.8

7.77

143

13.17

13.20

0.5G

Inspires air.

11.7

7.78

182

13.21

0.54

11.2

185

13.25

0.57

12.9

131

The anesthesia in this case was probably rather deep, and
consequently the animal had a relatively low ventilation and
high alveolar carbon dioxide pressure. There seems to be a ten-
dency for ventilation to increase as time goes on, probably as a
result of a gradual fall of the depth of anesthesia. When oxygen
is given before denervation, a small decrease in ventilation is

EFFECT OF CAROTID SIKUS DENERVATION. .9.0

observed during the next feiv minutes, indicating that some oxy-
gen lack existe. A compensation takes place later on. Denerva-
tion gives the ty^iical effect; ventilation decreases by about 27
per cent, and alveolar carbon dioxide rises by about 20 per cent.
Oxygen is now without effect. In this case — as in our earlier
control experiments — the effect of denervation is quite ob\dous,
even if the animal has been breathing oxj’-gen, so that complete
saturation of the hemoglobin is ensured. The conclusion is
that the decrease in ventilation after denervation is due to
the loss of the chemical stimulation of the carotid body by the
normal carbon dioxide tension of the blood. The blood pressure
rose immediately after the denervation but soon fell again to
the old level. A changed blood pressure cannot therefore be
responsible for the changed respiration.

Similar results have been obtained with an alveolar carbon
dioxide pressure somewhat below the physiological level, as il-
lustrated by the following experiment.

21. 2. 1940. Cat, I.f) kg, receives under ether anesthesia 35 ml
of a 1 per cent solution of chloralose and 4 ml of a 20 per cent
solution of urethane intravenously. Preparation as above.

Time

Ventilation

1. per min.

Respira-
tioii rate
per min.

Alveolar
carbon
dioxide
per cent

Blood
pressure
mm Hg

Arterial blood gases

0. per CO. per
cent cent

1,20

20.2

4.S8

190

17.1 36.2

(89.5 % sat.)

Inspires oxygen.
1.10

19.7

4.SS

192

20.5 33.6

(100 % sat.)

12'

15'

Inspires air.

4.01

18' 7 per cent oxygen in nitrogen gives considerable rise in ventilation.
30' Both sinuses clcncrvntcd. 7 ])cr cent oxygen in nitrogen gives a small
increase in ventilation,

35' Inspires air. 3.22

37' Inspires oxygen.

39' 0.S7 17.3 5.00 183

42' Inspires air.

45' O.SO 17.1 5.17 184

The decrease in ventilation after denervation is of the usual
order, though the increase in alveolar carbon dioxide is relatively
small. It. is possible that this was due to the fact that denerva-
tion was not absolutely complete. As in the case of the former

100 U. S. V. EDLER AND 6. LILJESTRAND.

Table

Dtcerehra-

Date

Weight

B 0

fore

d e

n e r V

a t i 0

■

Breathing air

Breathing oxygen

Ventila-
tion 1. pr.
min.

Eesp.

ratepr.

min.

Alveol.

COj

perc.

Blood

press.

mmHg

Eesp.
rate pr,
min.

Alveol.

COj

perc.

Blood

press.

mmHg

3/4 40

2.4

1.86

34.5

0.98

28.8

4.01

—

1.06

30.0

—

Vb 40

2.6

1.85

37.2

3.68

1.70

36.2

3.72

‘1.18

13.0

4.12

—

3/8 40

2.7

1.17

30.8

3.88

112

1.18

30.0

4.02

114

—

1.14

30.0

3.74

112

experiment, this also demonstrates that the effect is obtained,
even if the animal is breathing oxygen. A rise in the blood pres-
sure can obviously not be regarded as a determining factor for
the decrease in ventilation in this case.

Decerebration has the advantage that anesthesia is rendered
unnecessary, but it certainly involves a very heavy traumatiza-
tion of the central nervous system, which will often greatly af-
fect respiration. Sometimes after decerebration we have observed
different signs of respiratory disturbances, such as periodical
breathing, which is more or less alleviated by oxygen inhalation,
gasps etc. Sometimes we have seen a continuous increase in res-
piration, which must be ascribed to some stimulus of unknown
origin. Thus the alveolar carbon dioxide may show very low
values. We have observed, e. g., 1.38 and even 0.8 per cent.
Obviously such animals are unsuitable for experiments of the
kind in question, which aim at testing the effect of denervation
of the sinuses. We have succeeded, however, in reaching results
with decerebrated cats where respiration goes on with sufficient
constancy for our purpose. Table 1 gives the results of three
such experiments.

The results agree with those obtained under anesthesia. Oxy-
gen inhalation before the denervation reduces ventilation in
every case, but there is a further decrease after denervation.
Alveolar carbon dioxide percentage then rises by 0.4— 0.9.

After vagotomy.

1 .

ied cats.

EFFECT OF CAKOTID SINUS DENERVATION.

101

A f t e

r d e

n c r V a

tion

Breathing air

Breathing oxygen

Yentila-

Kcsp.

Alveot.

Blood

Yentila-

Besp.

Alveol.

Blood

tion ]. pr.

rate pr.

CO,

press.

tion 1. pr.

rate pr.

CO,

press.

mm.

min.

perc.

mm Hg

min.

perc.

mm Hg

0.90

31.2

5.09

0.9 0

30.0

4.92

—

‘0.86

l.i.

4.98

—

1.05

S2.0

4.13

Schmidt, Dumke and Dbipps (1939) have pointed out that
dogs and cate might react differently to denervation of the si-
nuses. Even though this is not very probable, it seemed desir-
able for us to perform some experiments -with dogs under the same
experimental conditions as those prevailing during the observa-
tions on cate. Our results are summarized in table 2.

In experiments 1, 2 and 3 no oxygen was given, and we can-
not, therefore, exclude an influence of relative oxygen lack.
There is a decrease in ventilation after denervation of 26, 11
and 14 per cent respectively. In the following experiments oxy-
gen was given at intervals, allowing us to compare the reactions
with, as well as without, oxygen. In experiment 5 the denerva-
tion was incomplete, as Avas shown by the reaction to 7 per cent
oxygen in nitrogen. This explains why oxygen is able to reduce
the ventilation, not only before, but also after the denervation,
Avhich otherwise does not occur. Both values, however, show a
small decrease (5 and 7 per cent) in relation to the correspond-
ing values before denervation, and a corresponding rise (6 per
cent) is observed in the alveolar carbon dioxide percentage.
In experiment 4 oxygen produced a very small reduction of
ventilation before denervation. "When this had been performed,
the ventilation was reduced by 14 and 13 per cent respectively,
and the alveolar carbon dioxide rose by 28 per cent. Experi-
ment 6 showed a great decrease. (24 %)Jnwentdation, when oxy-
gen was given, but there ,Ava’s^a;;furth^* 'redugtib!^(12 % of the

102 U. S. V. EULER AND G. LILJESTRAND.

Table

Anestheti-

B e

fore

d e

n e r V

a t i 0

1 Exp.
nr.

Weight

Breathing .lir

Breathing

oxygen |

Date

Ventila-
tion 1 . pr.
min.

Resp.

ratepr.

min.

Alveol.

CO,

perc.

Blood

press.

mmllg

Ventila-
tion 1 . pr.
min.

Resp.

ratepr.

min.

Alveol.

CO 3

perc.

: 1938

\ 1 ‘«/ii

■13.5

7.27

19.4

! 2

7.43

22.8

—

13 »/is

12.5

3-76

23.0

4.56

lot

—

1940

'Vs

13.5

5.14

22.0

4.03

5.06

22.0

4.01

140

13.5

—

2.72

14.0

6.44

112

}

3.57

15.9

• —

114

— -

—

|6 Ve

10.20

22.8

4.61

134

—

— -

—

7.86

18.7

4.78

120

* -r If

/ ‘Vt

7.57

34.0

5.34

132

—

5.05

26.9

5.53

134

8.77

37.0

—

137

—

8.12

36.5

4.76

140

—

7.95

39.6

4.00

134

original value) when denervation was performed. The corres-
ponding rise in the alveolar carbon dioxide was, in both cases,
much smaller than one would expect from the effect on ventila-
tion. In experiment 7 oxygen given at the beginning of the ex-
periment caused a very pronounced reduction in ventilation;
later on ventilation seemed to increase spontaneously, and sim-
ultaneously the usual effect of oxygen — a reduction of venti-
lation — was to a great extent reduced. Still the denervation
had the usual effect, the reduction in ventilation being 26 and
the increase in alveolar carbon dioxide 21 per cent.

The rise in the blood pressure was moderate in all the experi-
ments where vagotomy was not performed, and it does not seem
very probable that this rise can explain the effect on respiration.

Our observations on dogs are in complete harmony with the
results obtained with cats. For both types of. animals we have

EFFECT OF CAROTID SIKUS DEKERVATION.

103

2 ,

zed dogs.

A f t e

r d c

n e r V a

t i 0 n

Breathing air

Breathing

oxygen

Ventila-

Rcjp.

Alvcol.

Blood

tion 1. pr.

rate pr.

CO,

press.

min.

[

min.

pcrc.

mm llg

0.35

11.5

6.CI

27.0

—

3.23

20.0

5.2C

103

—

i 4.40

{

o.-) n
-..0

4.40

22.0

5.14

176

—

-2.53

13.G

G.85

144

' ' 3.39

15.9

6.75

150

—

— ■

}

—

6.58

15.9

5.00

152

—

■ —

—

;

—

i —

O.SG

30.0

5.70

172

been able to demonstrate, in confirmation of earlier experiments
on cats, that denervation of tlie sinuses leads to a decreased
ventilation and an increased carbon dioxide percentage of the
alveolar air.

Summary.

After denervation of the sinuses, a decrease in ^'entilatiou
and an increase in the alveolar carbon dioxide percentage was
observed in cats and dogs. The effect was reduced but not abol-
ished, if the animals were breatliing o.xygen. Neither anesthesia
nor increase of the general blood pressure can explain the effect.
It is concluded that the carbon dioxide tension of the blood under

* Both vagi cut.

- Denervation incomplete.

104

U. S. V. EULER AND G. LIUESTRAND.

plijrsiological conditions stimulates respiration, not only by a
direct action on tlie respiratory centre, but also reflexly over
the sinus mechanism.

Eeferences.

V. Euler, TJ. S., and G. Liljestband, Skand. Arch. Phy.siol. 1935.
71. 73.

— , — , Ibidem 1936. 74. 101.

V. Euler, U. S., G. Liljestrand and Y. Zotterman, Ibidem 1939.
83.. 132.

Gesell, B., and J. Lapides, Proc. 16th Int. Physiol. Congr. (Zurich)

1938. 1. 45.

Hahn, W., Arch. Kreislaufforsch. 1940. 6. 97.

Heyjians, C., and J. J. Bouckaert, J. Physiol. 1930. 69. 254.

— , — , Ergebn. Physiol. 1939. 41. 28.

Heymans, C., j. j. Bouckaert and L. Dautrebande, Arch. int.
Pharmacodyn. 1930. 39. 400.

Mertens, 0. and H. Bein, Pfliig. Arch. ges. Physiol. 1938. 241. 402.
Boos, J., and C. Eomijn, Arch, neerl. Physiol. 1937. 22. 233.
Saiiaan, a., and G. Stella, J. Physiol. 1935. 85. 309.

Schmidt, C. E., and J. Comroe, Physiol. Bev. 1940. 20. 115.
Schmidt, C. E., H. J. Comroe and B. D. Dripps, Proc. Soc. exp. Biol.,
N. y. 1939. 42. 31.

Schmidt, C. E., P. B. Dujike and B. D. Dripps, Amer. J. Physiol.

1939. 128. 1.

Selladurai, S., and S. Wright, Quart. J. exp. Physiol. 1932. 22.
285.

Witt, D. B., L. N. Katz and L. Kohn, Amer. J. Physiol. 1934, 107.
213.

From the Biochemical Department, Karolinska Institutet, Stockholm.

The Interaction of Amino Nitrogen and Carho-

liydrates.^

Bj-

GUNNAR AGREN.®

During the investigation of certain reactions in cell-free liver-
cxtracts (Agrex 1940) an interaction of carbohydrates with free
amino groups was observed. When reaction had ceased the de-
crease in the vax SLYKE-determinable nitrogen was not accom-
panied by a decrease in titrable amino nitrogen; neither was
any loss in total nitrogen or of ammonia noted even with an
80 % disappearance of yak Slyke nitrogen.

This great reactivity by incubation at 37° and pH 7.4 without
any loss in ammonia was rather surprising when compared with
earlier observations in model experiments with amino acids and
sugars, Maillard (1912) and later Borsook and Wasteneys
(1925), Euler and others (1926) showed that the interaction
between glucose and amino acids as alanine and glycine pro-
ceeded at a slow rate and to a limited extent at temperatures
between 20° and 40° and pH 7. In alkaline solution the rate of
reaction rapidly increased "odth the rise of temperature and pH.
During this process ammonia was produced. Borsook and
Wastekeys in accordance with Fearok and Moktgomery
(1924) in explaining the effect of the aldehyde group suggested
a disturbance of the equilibrium of the amino group by its prelim-
inary combination with the aldehyde group. Neuberg and
others (1927) noticed a speedy reaction at pH 7 between keto-
sugars (fructose, hexose-diphosphate) and amino acids as ala-

* Received 7 June 1940.

* Fellow of the Rockefeller Foundation 1938 — 1939.

8 — iOI323. Acta phys. Scandhiav. Vol. I.

106

QUNNAE AGREN.

nine, aspartic acid and glutamic acid, Th.e rapid interaction
between keto-sugars and amino groups did not produce any
further effect than that of the same amino acids and aldehydes.

[For further investigation regarding the reaction in the liver
extracts the model experiments were extended to the usual amino
acids and carbohydrates incubated with phosphate to a con-
centration observed in the liver extracts (Agren 1940).

Experiment.

Solutions of amino acids and sugars -were made in glass-dis-
tilled water containing glucose in a concentration of 0.5 M.
Phosphate was added in a mixture of Sorensen’s phosphate
buffers pH 7.4 to the neutralized amino acid-sugar solutions.
The phosphate concentrations Were M/15 and M/30 in two series
of experiments. The concentration of easily soluble amino acids
of 0.3 M or of practically saturated solutions was used. The pH
level was maintained at 7.4 (glass-electrode) during all experi-
ments. The amino nitrogen was followed with the van Slyke
method in 24 hour periods. The solutions were kept at N^-
atmosphere in Thunberg tubes and incubated at 40° or sealed
in tubes at 100°, Eesults of typical experiments are given in
table 1.

A comparison of the free amino nitrogen values shows, in agree-
ment with earlier observations, that the interaction at pH 7
between amino acids and glucose only takes place to a very
limited extent. The reaction is accelerated by raising the tem-
perature to 100° and increasing the phosphate concentration
from m/30 to m/15. There seems to be an optimal concentration
corresponding to about m/15. Raising the phosphate concent-
ration to about m/2 did not increase the rate or the extent of the
reaction. There is one especially reactive a min o acid: cysteine.
About 80 per cent of the free amino nitrogen in cysteine disap-
peared in 48 hours. In the experiment at 100° the well known
instability of cysteine at this temperature interfered with the
expected result.

The catalyzing effect of phosphate in the experiments was not
very marked but evident and had already been proved in the
liver extracts (Agren 1940). In other experiments the influence
of varying phosphate concentrations in nitrogen and oxygen

AMINO NITROGEN AND CARBOHYDRATES.

107

Table 1.

Loss of van StyX'c, nHrogen in the interaction between amino acids

and glucose.

Amino acid

' Phos-
j phato
( conccn-
j tration
in mol.

Temper-

nturc

Loss of van Slyke nitrogen
in per cent of zero value
after

ill hours 48 hours 72 honrsi

Total
loss in
72 hours,
per cent

Alanine 0.3 M. . .

i Glvcine 0.3 M. .

j Serine O.S 31. . . .

I Leucine 0.2 31. . .

j Valine 0.3 31. . . .
Cysteine O.S 3t. . .
Glutamic acid 0.2 31.
Arginine 0.3 M. . .
Histidine 0.3 31. , .
Tryptophan O.08 31.
Lysine O.S 31. . . .

31/30

3L15

31/15

31/30

31'15

31/15

3L30

31T5

31/15

31/30

31/15

31/30

31/15

31/30

31/15

31/30

31/15

31/30

31/15

31/30

31/15

31/30

31/15

31/30

31/15

40 0
40 C
100 C

40 C
40 C
100 C

40
40
100

100

100 C

40 0
40 C
100 C

40 C
40 C
100 C

40 0
40 C
100 C

40 C
40 C
100 C

100

10/30 minutes

8/30 minutes

15/30 minutes

8
9

12/30 minutes

15/30 minutes

35/30 minutes

21/30 minutes

15/30 minutes

21/30 minutes

0
0

0/30 minutes

21/30 minutes

15
13

15
15

atmosphere was further investigated. The experimental condi
tions were the same as those described above. A typical series

is given in table 2. 1,1 i, 4.1 •

The catalyzing effect of phosphate is demonstrable both in

aerobic and anaerobic conditions. In the phosphate-free solu-

108

GUNNAR AGEES’.

TftWe 2.

Disappearance of van Slylce nitrogen in aerobic and anaerobic alanine-
glucose solutions incubated with phosphate at JfO 0 and pH l.i

Sample

Van Slyke nitrogen in mg/cc after

0 hours

24 hours

48 hours

72 hours

Oj-l-PO^

4.20

3.98

3.80

3.81

0 .

4.20

4.15

4.10

4.12

N, -h PO 4

4.20

4.05

3.90

3.88

4.21

4.18

4.12

4.11

Concentration of alanine, glncose and phosphate respectively 0.3 M., 0.5 M.
and 0.07 M. Oj = oxygen, N, = nitrogen, PO^ = posphate.

tions tlie reaction "was almost unnoticeable. As a comparison
may be mentioned that a pbospbate-free solution of arginine
and glucose (0.3 M solution) kept at 10° and pH 10 sho’wed a 28
per cent loss of van Slyke nitrogen in 60 hours.

In alanyl-glycine the pKb is about 10 times as high as in ala-
nine. This difference in dissociation constant could eventually
influence the reaction -with the aldehyde groups. A 0.3 M solu-
tion of a racemic peptide preparation vas incubated vith glu-
cose and phosphate at pH 7.4 to a final concentration of 0.5 M
respectively M/25. The experiments v'ere performed in Thtjn-
BEEG tubes vdth oxygen atmosphere. The results of a typical
series are given in table 3.

Table 3.

Decrease of van Slyhe nitrogen in aerobic alanyl-glycine sugar solu-
tions incubated with phosphate at pH 7.4 and IfO C.

Sample

Van Slyke nitrogen in mg/cc after

0 hours

24 hours

48 hours

Phosphate

4.16

2.85

2.50

Control without phosphate ....

4.18

4.11

4.05

Concentration of alanyl-glycine, glncose and phosphate respectively 0.3 3L
0.5 M. and 0.04 II.

AMINO NITROGEN AND CARBOnYDRATES. 109

An increase in tlie dissociation of the amino group did not
accelerate the reaction ^vith aldehyde groups. The stimulating
effect of phosphate ions was more obvious in these experiments
than in the alanine-sugar experiments. Whether the phosphate
is influencing the amino group-sugar association or indirectly
stimulating the process, way of the carboxjdic groups, will
be further investigated.

In some experiments the tendency to an association between
amino groups and Iceto-sugars AVas examined. In the first series
alanine was incubated with fructose and phosphate at 40° and
pH 7.4, the final concentrations being respectively 0.3 M, 0.3 M
and 0.1 31. There was no provable disappearance of van Slyke
nitrogen in 24 hours. Negative results Avere also obtained in
other series with cysteine, cystine and alanyl-glycine. In this
connection it should be emphasized that there was no difference
in the reaction betAVeen natural and unnatural amino acids with
aldehyde and keto sugars. The negative results with cysteine
were surprising in aucav of the rapid and complete reaction of
that amino acid with glucose. Next methylglyoxal as represen-
tatiA’c of ketoaldehydes was used. This was fmmd to react in-
stantaneously. The results obtained in the first van Slyke de-
termination immediately after solving the amino acids in the
solution of methylglyoxal were not decreased in the next 48
hours, as demonstrated in table 4.

Tnlilo 4.

Disappearance of van Slphe nitrogen in aerobic metliylglyoxal-amino
acid solutions incubated at 0 with phosphate. pH = 7.4.

Amino acid

Van Slyke nitrogen in mg/cc after

0 hours

24 lours

Calc.

Found

Alanine

2.80

2.60

2.40

Alanyl-glycinc

2.80

2.40

2.89

Cysteine

2.80

0.66

O.60

Concentration of amino acids, metliyJglj’oxal and phospate respectively 0.2 M.,
0.2 M., 0,04 M.

It was obvious from all experiments that the equilibrium in
the reaction between amino and aldehyde groups at pH 7 was

110

GUNNAR IgREN.

with, one exception very unfavourable from a preparatory point
of view. Conclusively interest was focussed on the reaction be-
tween cysteine and aldehydes and eventually on the ketoalde-
hydes. In a solution of cysteine and glucose (0.3 M respectively
0.5 M) the reaction ceased when about 80 per cent of the van
Slyke nitrogen had disappeared. The results obtained then
remained constant for weeks. When the molecular ratio cysteine

—i— 0.1 M. Phosphate
0.1 M. barbituric

acid

Activating effect of phosphate on the cystein sugar reactions; pH 7.4, 40° C.
Concentration of cysteine and glucose respectively 0.3 M. and 0.5 M.

to sugar was 3 : 1 the van Slyke nitrogen disappeared to a point
corresponding to a complete reaction between sugar and cysteine.
A surplus of cysteine could clearly force the reaction to com-
pletion but was, it seemed, an uneconomical way of preparation
and afterwards equimolecular amounts of sugar and cysteine
were used. It is a Well known fact that the usual cysteine pre-
parations contain small amoimts of copper, which in the pres-
ence of air wiU oxidize cysteine to cystine. This was undesir-
able since experiments already had shown that cystine reacted
in the same limited way as the other amino acids. In all cysteme-

AMINO NITROGEN AND CARBOHYDRATES.

Ill

sugar experiments glass-distilled nrater was used and eventually
interfering copper was removed with diethyl-dithio-carbamide
(Agner 1939). Thunberg tubes and nitrogen atmosphere ex-
cluded any oxygen influence. Determinations of cysteine and
cystine were made according to the method of Kassel and Brand
(1938) readings being made from standard curves of cystine and

Figure 2.

Disappearance of van Slyke nitrogen in cysteine-sugar solutions; pH constantly
kept at pH 7.4 by adding 10 N HaOH. 40“ C.

Concentration of cysteine and glucose respectively 0.3 M. and 0.5 M.

cysteine preparations used in the experiments. Glucose was de-
termined with the orcinol method according to the modification
of Sorensen and Haugaard (1933), and phosphate according
to Ldndsteen and Vermehren (1936). The optimal conditions
of the reaction were studied in series of experiments. When the
cysteine-glucose reaction was followed by van Sdyke deter-
minations in 0.1 M solutions of phosphate and barbituric acid
buffers (pH 7.5) the reaction became more definite in phosphate

112

GDNNAR AgREN-.

solutions. (Figure 1.) Th,e pH at the end of the experiment -was
in hoth solutions about 5.8.

In two series of experiments solutions of cysteine and sugar
were constantly kept at pH 7.4 by adding 10 N NaOH with due
precaution being taken to exclude air. Phosphate was present
in one of the series in a final concentration of 0.1 M. The de-
crease in VAN Slyke nitrogen is pictured in figure 2. The activ-
ating effect of phosphate ions was manifest.

Pigure 3.

Disappearance of van Slyke nitrogen in anaerobic cysteine-glucose solutions
incubated at 40° C. in 0.2 M. buffer solutions. Reaction time 24 hours. The values
calculated as decrease in nitrogen, mg/cc in 24 hours.

Citrate buffer pH 2.6, acetate pH 3.6 and 4.2, phosphate pH 5.5, 6.1 and 7.i.
Concentration of cysteine and glucose respectively 0.3 M. and 0.5 M.

In contrast to the other amino acids the cysteine amino group-
aldehyde reaction was demonstrable in acid solutions. Figure 3
shows the rate of reaction (disappearance of van Slyke nitro-
gen) in 0.2 M buffer solutions between pH 2.6 and 7.2. The op-
timal pH of the reaction was not in acid solutions. The spontan-
eous oxidation of cysteine by air starts at pH 8 and increases
with the rise in pH. Accordingly the cysteine-glucose reactions
were studied at pH 7.4.

AMIKO NITROGEN AND CARBOHYDRATES.

113

In control solutions vritli cysteine alone the cysteine concentra-
tion had not changed after 48 hours. The stability of the cysteine-
sugar product was rather strong in acid solutions. Shaking for
30 minutes in the van Slyke determination did not split the
product. "When mixed with HCl to a final concentration of 10
per cent the cysteine-sugar product was not split after remaining
8 hours at room temperature. Heating in a sealed tube in water
bath for 30 minutes in 10 per cent HCl hydrolysed 10 per cent
of the compound.

A series of equimolecular cysteine-glucose solutions (0.3 M)
were constantly kept at pH 7.4 and 40° 10 N NaOH being added
from a micro burett while N, was bubbling through the solution.
Parallel determination of van Slyke nitrogen, acetone titra-
tions on amino nitrogen and cysteine-cystine determinations
were made. The results are given in table 5.

Table 5.

Parallel determinations of van Slylce nitrogen, iitrahle amino groups,
cysteine and cystine in cysteine-glucose solution incubated at 0

and pH 7.4.

Von
Slyke
nitrogen
in mg/cc

CO n/40
HCl re-
quired in
titration
of 75
c. mm. of
solution

c. mm. 10
NaOH
added
per cc
solution
to bring
pH to 7.4

Cysteine
in mg/cc

Cystine
in mg/cc

cc n/40
NaOH
added
from time
0 calc, on
75 c. mm.
solution

Column 3
— blank
values —
column 7
= titrable
amino
groups

3.95

1.32

27.5

0.69

3.18

1.38

1.95

28.0

0.060

0.69

2.7G

1.41

1.67

28.0

0.112

0.67

2.56

1.43

2.81

27.5

0.180

0.64

2.00

1.48

2.00

23.6

0.228

0.63

1.76

1.67

1.83

21.1

0.284

0.66

1.00

1.60

2.00

13.5

0.344

0.63

Concentration -of cysteine HCl and sugar respeoHvely 0.3 M. and 0.5 M.

It is made clear in column 8 (the titration values corrected
for blank values by the titration and the added amounts of HaOH)
that the nitrogen atom in the compotmd formed can still be ti-
trated as amino nitrogen in spite of the apparently decreased
basic properties of the cysteine-sugar compound. In all prob-

114

GDNNAll AQREN.

ability tbis must contain tbe sulpbur atom in sucb a position
as to be underterminable either as cysteine or cystine by tbe
method here used. These conclusions are drawn from the corres-
ponding amounts of amino nitrogen and cysteine left after the re-
action has ceased.

Various efforts were made to isolate the cysteine-sugar com-
pound. In this connection may be mentioned that Schubert
(1936) studied reactions between organic sulphur and aldehyde
compormds. Fractionation with alcohol, phenol and pyridine
gave inhomogeneous products. There seemed to be a tendency
in alcohol-water solutions of this product to dissociate into sugar
and cystine. Solutions containing the cysteine-sugar product
were aerated with traces of FeCIj to remove cysteine which had
not reacted with sugar. Then the cysteine-sugar compound was
precipitated with CuAcj and ethyl alcohol to 60 per cent. The
precipitate was treated with HjS which was removed with bubb-
ling Nj. Through this oxidation and reduction a product was
obtained containing no free van Slyke nitrogen but with sul-
phur clearly determinable as cysteine-sulphur according to EAs-
SEL and Brand (1938). A similar type of substance was obtained
directly when cysteine reacted with methylglyoxal in equi-
molar amounts (table 4). It is of special interest to note that the
thiol test with sodium nitroprusside gave a negative result on
the same conditions. While it Was possible to isolate this lateral
reaction with a free thiol group, the sulphur in the main reac-
tion was fixed in a position of another type, where it could not
be determined as thiol or S — S sulphur.

The method finally adopted to isolate the cysteine-glucose
compound was as follows. 30 mM each of cysteine and glucose
were solved in 50 cc glass destilled water at pH 7.4 and 60 mg.
diethyl-dithio-carbamide added. After gentle shaking the die-
thyl-dithio-carbamide-copper was extracted in five minutes with
amyl alcohol. The water solution was kept at 40° C and pH 7.4
as described above. The reaction usually ceased after 12 hours
(fig. 2) and the solution was concentrated in vacuum to a sirup
and precipitated with water-free alcohol 10 — 16 volumes. The
mixture was kept over night at 0°, centrifuged and the precip-
itate dried with alcohol and ether. 2 g. of the precipitate was
thoroughly extracted with 20 cc of concentrated acetic acid.
About 50 per cent of the material was solved in the acetic acid,
and the solution was precipitated with ether. Both fractions

AMINO NITROGEN AND OABBOHYDRATES.

115

were dried with alcohol and ether. Analysis of the soluble and
insoluble acetic acid fractions ate given in table 6. Sulphur was
determined with the Pregl method.

Table 6,

Analysis of cysteine-ghtcose compound.

Sample

Total

nitrogen

Sulphur

Van

Slyko

nitrogen

Cysteine
+ cystine

Sugar

Yield

Calculated for

per cent

CjHijOjNS

4.9

11.3

Insoluble in acetic acid

4.8

11.2

Soluble in acetic acid .

2.2

5.8

1.8

The acetic acid extraction removed small quantities of sugar
and cysteine, which had not reacted, together "v^ith a part of the
final product. The rest of the material was obtained in a practi-
cally pure form. It was easily soluble in water. "^Tien exposed
to air it absorbed Water and it was necessary for the sake of anal-
ysis to dry the product to constant weight over PsOj. The con-
stitutional similarity with the thiazol compounds (see discussion)
made it desirable to test the effect of oxidation reactions on the
compoimd. Weak oxidants, as the oxygen of air, had no activ-
ity even in the presence of copper. Hydrogen peroxide easily
split the product into cystine and sugar. The same effect was
obtained by dehydration. Methylenblue was reduced and free
cystine and glucose isolated. The ultimate biological effect was
tested with the phycomyces method. This test was performed
by dr Nielsen of the Biological Department of Carlsberg Labor-
atory. The thiazol derivative did not increase the activating
influence of the pyridine part of vitamin B^. It is quite clear
that the thiazol compound could not be coupled directly in its
isolated form.

Discussion.

The investigation of the reaction between aldehyde groups
and amino acids was undertaken with a view to isolating some
of the reaction products for a further study of the reactivity of
the linkage between the amino and aldehyde groups. The ten-

116

GUNNAR AQREN.

dency to form bonds at tbe physiological hydrogen concentration
is very -weak in all amino acids ■with the exception of cysteine.
In this case the aldehyde group is Hnked to t'Wo groups in the
molecule of the animo acid, the thiol and amino group. This
gives strength to the compound. A comparison of the reactivity
of amino acids and dipeptides proved that shght increases in the
dissociation of the amino groups did not facilitate the reaction.
From a structural point of view it may be mentioned that there
Was no difference in the reactivity of the natural and unnatural
amino acids. Experiments later to be published confirm the
•view that the undissociated amino groups are the most reactive.
The activating influence of phosphate remains to be explained.
Several possibilities offer themselves. The phosphate ions may
act directly on the aldehyde-amino group association; it is known
that metaphosphoric acid derivates form compounds with amino
groups. It has also been demonstrated at Warburgs laboratory
that the phosphoric acid forms compounds with the carboxyhc
groups in intermediately formed carbohydrate compounds of the
fermentation reaction. Experiments now in progress seem to
show that the reactivity of alanine with glucose is stimulated
by the estrification of the carboxylic group. It is possible that
the phosphate ions may act in some similar way.

Interest was concentrated on the cysteine sugar products. It
is possible to prove that complete reaction depends on the form-
ing of a second bond to the thiol group. A comparison between
the rate of disappearance of van Slyke nitrogen and thiol group
showed that the initial reaction included the forming of the bond
between the amino and aldehyde groups. The properties of the
finally isolated compound corresponded to a thiazol compound.
Structural similarities -with the thiazol part of the vitamin Bi
made a testing of the biological properties desirable. Obviously
the structural differences could not be eliminated by the syn-
thetical activity of the phycomyces bacteries.

CH CH

S -fNH S -j-N-CHo-E,

H. • ^ C - COOH OH • CH, • CH, • i_=C • CH3

The cysteine-glucose compound. The thiazole half of vitamin Bi.

ASimO NITROGEN AND CARBOHYDRATES.

117

Several efforts -were made to oxidize or dehydrate the isolated
compound hut the result was unexceptionaUy an opening of the
thiazol ring. It is possible that by methylation of the amino
group, decarboxylation or estrification of the carboxylic group
and by replacing the glucose radical with some phospho-glycerin-
aldehyde compound a more active thiazol compound can be
obtained. From a biological point of view this reaction between
sulphur containing amino acids and aldehydes may be of further
interest. Anyhow the occurrence of considerable amounts of
cysteine in the liver extracts may explain some of the rapid and
complete disappearances of van Slyke nitrogen in these extracts
{Agrbn 1940).

Summary.

1. The reaction between glucose and the usually occurring
amino acids and some peptides has been studied at pH 7.4.

2. Only with cysteine an approach to a complete reaction
was obtained.

3. A cysteine-sugar compound with thiazol structure has
been isolated.

4. The reaction between amino acids and glucose is acceler-
ated in the presence of phosphate.

The author wishes to express his gratitude to Professor K.
Linderstrom-Lang for extending the facihties of the Carlsberg
Laboratory and for his untiring interest in the work. This has
been supported by a grant from the Therese and Johan Anders-
sons Minne Poundation.

Eeferences.

Agxer, K., Naturwissenschaften. 1939, 27, 418.

Bobsook, H. and H. Wastexeys, Biochem. J. 1925, 19 , 1128.

V. Euler, H. and E. Brunius, Hoppe-Seyl. Z. 1926, 161 , 265.
Fearox, W. R. and E. G. Moxtgommery, Biochem. J. 1924, 18 , 576.
Kassel, B. and E. Brand, J. Biol. Chem. 1938, 125 , 117.
Luxdsteex, E. and E. Versiehrex, C. R. Lab. Carlsbere. Serie
chim. 1936, 21 , 147.

Maillard, L. C., C. R. Soc. Biol., Paris. 1912, 154 , 66.

118

GUNNAR AGREN.

Neuberg, C. and M. Kobel, Biochem. Z. 1927, 182, 273.

Schubert, M. P., J. Biol. Chem. 1936, lU, 341.

Sorensen, M. and 6. Haugaard, C. E. Lab. Garlsberg Serie chim.
1933, 19, nr 12.

Agren, G., Ibid. 1910, 23, 173.

(From tbo Biochemical Department, Karolinska Institutet, Stockholm.)

On tlic Peptidase Activity in the Cattle Muscle.^

GUNNAR AGREN.*

To fneilitate an e.vaminntion of transamination from peptides
in cattle muscle, it was desirable to separate the reaction from the
interfering influence of eventually contaminating peptidase activ-
ity. Tlie prc.sent paper is confined to a study of peptidase activity
in the cattle muscle. The experiments were extended in two di-
rections. To begin with, an effort was made to suppress the pep-
tidase activity by means of the usual enzyme inhibitors. When
this method proved unsatisfactory several peptides were investi-
gated with regard to their capacity to resist enzymatical digestion
Under the condition.s maintained by Bkauxstkin and Kritzmann
(Jf>37) in their study of transamination, glycyl-aminobenzoic acid
and valyl-glycine was not split.

Expcrimentnl Procedures.

Cattle diaphragm muscle was chilled in the slaughter-house
immediately after the death of the animal, and cut very fine in a
mincer. The minced muscle was suspended in 4 parts of 0.125 per
cent IvHCOj. A part of the solution was directly taken to pep-
tidase analy.sis and the rest was incubated in TiiUNnERO tubes at
40”, and shaken for 30 minutes for a new test. Within this space
of time the transamination would have ceased. There was a possi-
bility that the peptidases in the muscle would be more completely
extracted during the shaking period. After the digestion there
followed a titration of the newly formed amino and carboxylic
groups with the micro methods of Linderstrom-Lang and Hol-

' Ileccived for pnblication 29 June 1940.

* Fellow of the Rockefeller Foundation 1938 — 1939.

120

GUNNAE IqEEN.

TER. The time for digestion was limited to 30 minutes according to
reasons stated above. Glutamic acid was determined by the Jokes
and Moeller method (1928) in the modification of Brauksteik
and Kritzmakk (1937).

Results.

Alanyl-glycine (AG) and alanyl-glycyl-glycine - (AGG) were
easily split by peptidases in the muscle extract as seen in table 1.

Table 1.

Digestion of alanyl-glycine (AG) and alanyl-glycyl-glycine (AGG) in
fresh and incubated muscle.

Fresh muscle +

Incubated muscle -h

AGG

Digestion in c.mm

3.2

11.4

5.9

12.5

Concentration of peptide eolutions 0.2 M. Digestion nvith 7 c.inm. of substrate
and 7 c.mm. of filtered mascle extract. Digestion values in c.mm. n/20 HClin
90 per cent alcohol. Incubation time 30 minutes. Temperature 40°.

Muscle tissue evidently contained considerable amounts of
dipeptidase and aminopolypeptidase. The enzyme concentrations
were increased by the incubation process. The undesirable effect
of the peptidase activity was clearly demonstrable in a transami-
nation experiment with alanyl-glycine, the experimental condi-
tions being the same as given by Braunsteik and Kritzmakk
(1937). Table 2.

Table 2.

Digestion of alanyl-glycine in muscle extract.

Sample

Incubation
time in
minutes

Total van Slyke
nitrogen in
mg/cc.

Van Slyke nitrogen
of the glutamic
acid fraction in
mg/cc.

Alanyl-glycine

0.76

0.38

0.98

0.83

Control

0.19

0.24

0.19

0.22

3 g. of muscle + 1.2 cc of 2 per cent KHCO 3 + 3 cc (90 mg) of a d,l-
alanyl-glycine solution + 8 cc of water. Control without alanyl-glycine. In-
cubation at 40°.

ox Tin: PBPTiDASE ACTivrrr ix the cattle muscle. 121

Tiie difference in van Sh-ke nitrogen between the alanyl-glycine
solution before incubation, and the control solution, corresponds
to an almost complete digestion of the added alanyl-glycine in 30
minutes. The increase in glutamic acid van Slyke nitrogen depends
on tlie presence of glycine, liberated during the dige.stion.

Next an effort was made to .supprcs.s the peptidase activity
by cnr.yme inhibit or.s. The choice of inhibitor wa-s to some extent
free, since Bn.AUxsTEix and Kiutzmaxx (1937) had demonstrated
that tran.^amination proceeded rather independantly of o.xygen or
nitrogen atmosphere. Some of the results obtained with enxyme
inhibitons are given in table 3.

, Table 3.

Jtiflvcncc of enzutne inhibiiors on the imisdr. (U-pcpUdascs.

I’cptidc added j

1 1

1 Inhibitor j

1 j

Digestion in c.mm.

; AlanvI-glycinc O.C M

1 S.S :

i Alanyl-glycinc 0.2 M

Dromo.arotatc

1 ;f)0(X)

8.2 1

i .Alanrl-glvcinn 0.2 M

Sodium nr.scnito
m/100 1

8.2 :

\ Al.anyl-glydnc 0.2 M i

Potassium rynnide

; 5.8 !

m'lOd

i !

■ Alaovl-pivcine 0.2 M

i * ' ;

Sodium fiimridc
m/100

8.1 1

Lcncyl-giycinc 0.2 M •

2.2

Sodium evanido i

2.8 '

m/100 j

Moscle citr.net incubated 30 rtiinntcs at 40’, filtered. Digestion with 7 c.mm.
of enhstratc + 7 c.inm. of muscle extract. Dige.stion ns c.mm. d/20 HCl in
90 per cent alcohol.

The results with the enzyme inhibitors were not promising,
as a cause of which interest turned to the .substrates. Peptides
were tested regarding their capacity to resist enzymatical di-
gestion. Glycyl-aminobenzoic acid is known to be slowly di-
gested by giycerinc extracts of tlie intestinal mucous membrane,
i'he peptide wa.s sjuithezised according to one of Fischer’s
methods (1905), the ortho-derivate being preferable to the para-
compounds for enzymatical analysis, as it was more easily
soluble.

9 — ’iOJ323. Acin a/ij/s. f/candtuav. Vol. /.

122

GUNKAB IGREN.

Solubility of glycyl-aminobenzoic acid compounds.
Solubility in mg/cc at 20°.

Ortho-derivate 11 mg

Para-derivate 5.5 mg

Meta-derivate 13 mg

In the enzymatic experiments the solution of the ortho-
compound was saturated at 40°, the concentration being about
0.1 M (19 mg/cc). Still the solubility of glycyl-aminobenzoic acid
was so low, that in functioning as amino group donator, the effect
was found to be just within the range of the analytical methods
used in the transamination experiments. A few other peptides
were included in the enzymatical analysis: Leucyl-glycyl-glycine,
leucyl-phenyl-amino-acetic acid and valyl-glycine. Valyl-glycine
was synthezised according to Fischer (1907). The results of the
digestion with muscle peptidases are listed in table 4. The bicar-
bonate suspension of muscle was incubated at 40° for 30 minutes,
filtered and used as enzyme solution.

Table 4.

Pepiide digestion in muscle extract.

Peptide added

Digestion
in c,mm.

Substrate + en-
zyme solution
in c.mm.

Alanyl-glycine 0.2 IL

8.8

7-1-7

Glycyl-aminobenzoic acid 0.1 M

7 + 7

Leucyl-glycyl-glycine 0.2 11

1.8

7 + 7

Leucyl-glycyl-glycine O.Ol 11

1.6

70 + 7

Leucyl phcnylaminoacetic acid 0.05 M. .

0.6

2,0 + 7

Valyl-glycine 0.2 11

0.2

7 + 7

The enzyme-substrate solntions incubated 30 minutes at 40’. Digestion in
c.mm. n/20 HCl in 90 per cent alcohoi. Glyeyl-aminubenzoic acid followed by
titrations on the carboxylic groups with n/20 (CHshNOH.

The only two peptides which could be used in transamination
were apparently glycyl-aminobenzoic acid and valyl-glycine. Test
experiments were made with the standard concentration of amino
group donators (100 [j, Mol/g of muscle) in the muscle-brei using
the two peptides as donators. An increase of van Slyke nitrogen

ox THE PEPTIDASE ACTIVITT IN THE CATTLE MUSCLE, 12B

indicating a liydrolysation. could not be proved after 30 minutes’
incubation at 40"*, as seen in table 5.

Table 5.

Digestion of peptides in tmtscle extracts.

Peptide added

Incubation
time in minutes

Van Slyke
nitrogen in
mg/cc.

alanvl-glydno

0.7S

0.96

j vaM-glycine

0.73

0.72

* glycyl-aminobenzoic acid

0.76

{

0.77

3 g. of muscle + 1.2 cc. of 2 per cent K.HCO 3 + peptide solution (100 g
MoHg muscle) + water to fill up to 12 cc. Shaking at 40’ in water bath.

Discussion.

The peptidase activity in cattle muscle is about half as high
as in the intestinal mucousa of hog, -when the digestion of alanyl-
glycine is taken as a reference. Especially di- and tripeptides
containing alanine are especially sensitive to attack. Replacing
alanine in alanyl-glycine with leucine or valine serves as an
effective protection for the peptide against digestion. It ^vas of
interest to note that the splitting of alanyl-glycine was to some
extent inhibited in the presence of cyanide, while the digestion
of leucyl-pep tides was not influenced. It is a well known fact
that the affinities of other enzyme systems, e. g. lactic dehydro-
genase, are not strictly limited to one substrate. Homologous
compounds may also react, but the rate of reaction rapidly falls
off as the carbon chain is lengthened. EoUowing the hint, it was
possible to demonstrate that glycyl-aminobenzoic acid and valyl-
glycine could be suitable amino group donators in transamination
experiments, as they were nor split by peptidases in the short
time taken for the experiment.

Summary.

Cattle diaphragm muscle contains di-peptidases and amino-
polypeptidases, splitting alanyl-peptides especially. Valyl-glycine

124

QUNNAR AGREN.

and glycyl-aminobenzoic acid are not hydrolized in experiments
of short durability. The digestion of alanyl-peptides is partially
inhibited by cyanide.

The author wishes to express his gratitude to Professor K. Lin-
deestrom-Lang for extending the facilities of the Carlsberg Labo-
ratory, and for his untiring interest in the work and valuable
criticism of the manuscript.

The work was supported by a grant from the Therese and Johan
Andersson’s Minne Foundation.

References.

Braunstein, a. E., and M. 6. Kritzmann, Enzymologia. 1937, 2 , 129.
Fischer, E., Liebigs Ann. 1905, 340 , 123.

— , Ibid. 1907, 354 , 12.

Jones, D., and 0. Moeller, J. Biol. Chem. 1928, 79 , 429.

Aus dem Physiologischen und Chemischen Inslitut der Veterinar-
hoehschule zu Stockholm.

Die StiilbDitat der organisclien PliospliorverTbin-
dimgen und Phosphatase in Pferdehliit hei
desseii Aiifbewahriing in yitro.^

Von

KNUT SJOBERG.

Es ist eine alte Erfahrung, dass eine Hydrolysierung der Phos-
pliorsaureester in hanaolysiertem oder mit Kohlensaure gesattig-
tem Blut stattfindet. Es finden sich jedoch keine eingekenderen
Untersuchungen uber die Stabilitat des Blutes, speziell der Pbos-
pborverbindungen, in vitro vor. Nacbdem EAhbaeus (1939)
und Bekgenhem (1939) gezeigt haben, dass die Erythrozyten,
besonders im Pferdeblut, aus der Miinzform in eine melir spba-
riscbe Form ubergehen, falls das Blut bei 38 — 40° C aufbewabrt
■wird, drangt sich die Frage auf, ob diese Veranderung mit einigen
bedeutenderen chemischen Umwandlungen vereinigt ist. Nach
der Ansicht der genannten Forscher beruht die Veranderung der
Form der Blutkorperchen auf der Bildung von Lysolecithin im
Plasma, einem Stoff, der die Blutkorpermembran beeinflusst.
Die Permeabilitat derselben n im mt zu, und Wasser nebst einem
Teil geloster Stoffe diffundiert aus dem Plasma in die Erythro-
zyten und eventuell auch in entgegengesetzter Richtung. Es lasst
sich denken, dass schon dieses Verhalten eine solche Anderung
des Milieus zu bedingen vermag, dass die Phosphoresterasen ihre
hydrolysierende Wirkung ausuben konnen.

Die vdchtigsten Phosphorsaureester in den Erythrozyten
bestehen aus Hexosemonophosphorsaureestern (Kay und Ro-
bison, 1924), Glyzerinsaurediphosphorsaureestern (Jost 1927)
und Adenylpyrophosphat (Adenosintriphosphorsaure) (Loh-

1 Der Redaktion am 1. Juli 1940 zugegangen.

126

KNUT SJSBERG.

MANN, 1928, 1929). Dazu kommen die phosphorhaltigeu Lipoide
vom Lecithintypus, die hier der Kiirze halber mit Lecithin he-
zeichnet werden. AUe diese Phosphorverbindungen konnen von
verschiedenen Enzymen gespalten werden, von denen im Blut
nur die Monophosphoresterase eingehender untersucht wurde.
Pyrophosphatase ist im Blut von Kay (1928) und von Eoche
(1931) nachgewiesen worden, wurde aber keinen eingehenderen
Untersuchungen unterzogen. Ich vermochte jedoch neulich nach-
zuweisen, dass dieses Enzym hauptsachlich in den Erythrozyten
vorkommt, und dass seine Menge bei verschiedenen Tierarten
variiert (noch nicht veroffentlicht).

Das Lecithin kann nach Belfanti, Contabdi und Ercoli
(1936) auf zweierlei Art nach folgendem Schema gespalten werden:

Lecithin V.

. /

Lecithinase A
/

ungesattigte Eettsaure
+ Lysolecithin

'^Lecithinase B

{ Fettsauren +

Cholinglycerinphosphorsaureester

Cholin -f- Glycerinphosphorsaureester
Glycerin -}" Phosphorsaure.

Die primar gebildeten Stoffe werden weiter von anderen Enzy-
men zerteilt, so dass schliesslich anorganische Orthophosphate
freiwerden. Das Vorkommen von Lecithinase A im Blut ist durch
Bergenhems (1939) Untersuchungen als sichergestellt zu betrach-
ten. Inwiefern sich Lecithinase B und diejenigen Enzyme, welche
weiter die Spaltungsprodukte hydrolysieren, im Blut vorfinden,
scheint nicht untersucht worden zu sein.

Der Zweck der vorliegenden Arbeit lag in der Untersuchung des
Verlaufes der Spaltung der Phosphorsaureester im Pferdeblut,
das teils bei Zimmertemperatur (etwa 17° C) aufbewahrt, teils auf
Korpertemperatur (37°) erwarmt und gehalten wurde.

Die Untersuchung ergab, dass die Phosphorsaureester bei Zim-
mertemperatur bedeutend langsamer hydrolysiert werden als bei
Korpertemperatur, ein Verhalten, das ja mit Kiicksicht auf die
Abhangigkeit der enzymatischen Eeaktionen von der Tempera-
tur zu erwarten war. Bei Zimmertemperatur zeigten die Blut-

PHOSPHORVERBINDTJNGJEN UKD PHOSPHATASE IM PFERDEBLUT, 127

proben wahrend der ersten 24 Stunden nur eine unbedeutende
Spaltung. Auch das nach der Hamatokritroetbode bestimnite
Volumen der Blutkorperchen anderte sick nicbt. AUmablich be-
gannen die Blutproben zu hamolysieren, ixnd im Zusammenbang
biermit trat eine Hydrolyse der Pbospborsanreester ein, deren
ScbnelUgkeit mit fortdauerndem Blutkorperzerfall zunabm.

In denjenigen Ballen, in welcben das Bint bei 37° C aufbewabrt
wurde, anderte sich die Form der Blutkorperchen in 24 Stunden
in Eicbtung einer Spbarozytose. Im Zusammenbang biermit
stiegen die Hamatokritwerte. Setzt man voraus, dass die Ober-
flache der Blutkorpermembran unverandert ist, und berecbnet
man den Grad der Anderung des Volumens, wenn die Form aus
der Miinzform in die sphariscbe iibergebt, so ergibt sicb eine
Anderung von 1 ; 1.75. In vorliegenden Versucben anderten sicb
die Hamatokritwerte nacb 24 Stunden durcbschnittlicb im Ver-
baltnis von 1 : 1.6, also zu einem etwas niedrigeren Wert als dem
berechneten. Es ist jedocb zu beacbten, dass nicbt alle Blutkor-
percben vor und nacb der Erwarmung gerade die ideale Miinzen-
bezw. Spbarenform besitzen. Ausserdem war zu'weilen eine
schwacbe Hamolyse inzwiscben eingetreten, -was eine Vermin-
derung des Hamatokritwertes bedingt.

Zur Untersucbung einer eventueU gesteigerten Permeabibtat
im Zusammenbang mit der Veranderung der Blutkorperchen wurde
die Menge geAvisser anorganiscber lonen im Plasma vor und nacb
der Erwarmung bestimmt. Es zeigte sicb, dass Na *, Ca * • und
Cr, aus dem Plasma zu den Erytbrozyten in derselben Proportion
wie das Wasser wanderten, weshalb die Blutkorpermembran fiir
diese lonen voUstandig permeabel worden war, K * wanderte in
entgegengesetzter Ricbtung von den Erytbrozyten zum Plasma,
was auf dem grosseren Gebalt der ersteren an dieser lonenart be-
rubt. In den bei 17° aufbewabrten Blutproben gescbab wabrend
der ersten 24 Stunden kein Austauscb von lonen zwiscben Plas-
ma und Blutkorperchen.

Die Untersucbung der Resistenz der Blutkorperchen in HaCl-
Losungen von wecbselnder Konzentration zeigte eine Vermin-
derung der Resistenz.

In den meisten Fallen liess sicb nacb 24 Stunden bei 37° keine
Hamolyse feststellen, nacb 48 Stunden fand sicb jedocb eine sebr
kraftige oder vollstandige Auflosung der Blutkorperchen vor.
Aucb in denjenigen Fallen, in welcben keine Hamolyse stattge-
funden batte, begannen die Pbospborsanreester zu zerfallen.

128

KNDT SJSbBEG.

tind nach. 2 — 3 Tagen wax die ganze Menge sog. sauxelosliclien
Phospliors in anorganiscli gebundenen Pboxspbor iibergegangen.
Dieses zeigt, dass das Blut alle die Typen von Phospbatasen ent-
halt, welcbe die Pbospborsaureester bydrolysieren. In gewissen
Fallen erbielt icb mebr anoxganiscben Pbospbor als es dem ur-
spriinglicben saureloslicben Pbospbor entspracb. Der tJber-
scbuss muss vom Lecitbin berstammen, das also einer Zerteilung
nacb dem genannten Schema anbeimfallt, wenngleicb diese Hy-
drolyse verbaltnismassig langsam verlauft.

Die Pbospbormonoesterase kommt beim Pferde sowobl im
Plasma als aucb in den Erytbrozyten vor. In den bier untersucbten
Fallen variierte die in BoDANSKY-Einbeiten (B. E.) ausgedruckte
Pbospbatase im Plasma zwiscben 2.1 und 6.1, durcbscbnittl. 3.1,
und in den Erytbrozyten zwiscben 9.5 und 16.4, durcbscbnittl.
12.0, alles pro 100 ml berecbnet. Der Gebalt an Pbospbormonoes-
terase war also in den Erytbrozyten ungefabr 4mal bober als im
Plasma.

Die Pyropbospbatasen kommen ebenfalls in bedeutend grosse-
rer Menge in den Erytbrozyten als im Plasma vox. Das Verbaltnis
ist nacb unveroffentlicbten Untersucbungen im Durcbscbnitt
24: 1.

Da die Pbospborsaureester in den Erytbrozyten vorkommen,
muss die Spaltung entweder daselbst oder aucb im Plasma vor
sicb geben, nacbdem die Ester in dasselbe binausdiffundierten.
Dieses kann jedoch erst nacb einer geniigenden Zunabme der
Permeabilitat der Blutkorpermembran stattfinden.

Der Umstand, dass eine kraftige Pbosphatasewirkung aucb bei
17° in bamolysiertem Blut eintreten kann, scbeint darauf binzu-
deuten, dass eine Aktivierung der Enzymtatigkeit bei der Hamo-
lyse stattfindet. Die Bestimmung der Phospbatasewirkung im
Blut gescbiebt in der Regel in bamolysierten Proben. In unver-
anderten Blutproben konnen sicb die Enzyme in einem solcben
Zustand befinden, dass sie keine Wirkung ausiiben. Man weiss
aus anderen Fallen, dass das Enzym erst in eine aktive Form
iibergefiibrt werden muss.

Zwecks Untersucbung des Einflusses der Hamolyse wurde die
Pbospbormonoesterase nacb Bodansky (1933) in einer mit dem
Blut isotoniscben Reaktionsmiscbung bestimmt. Obgleicb die
Erytbrozyten nicbt der Hamolyse anbeimfielen, erbielt icb fiir die
Pbospbatasewirkung denselben Wert wie bei der gewobnbcben
Bestimmung in bamolysierten Proben. Die Pbospbormonoeste-

PnOSPHOBVERBINDtTKGEN UNO PHOSPHATASE IM PFERDEBLUT. 129

rase sclioint sich an der OberflSiche des Blutkorpcrcliens zu be-
finden, wo sic die Glyzorinphosphorsiiure in der Losung bydro-
lysiercn kann, abor nicbt die Phospborsaureester in den Blutkor-
pcrcben.

Diesc Ecsultate dcuten darauf bin, dass sicb die Pbospbormono-
esterasc nnd die Pliospborsiinxeestor in den Blutkorpercben in
oiuein soicben Zustand bcfinden, dass die ersteren nicbt die letz-
teren nnzugreifcn vcrmogcn. Erst nacbdem die Blutkorpercben
gowisscn Yerandcrungen anbcimgefallen sind, beginnt das Enzyni
seine Tutigkcit zu entfalten. Evcntucll kann hierbei irgendein
Aktivator bcteiligt- soin.

In kiinstlicb luimolpiorton Bintproben konnte man auf Grund
der ftubcren Hinwcise eine relativ scbncllc Hydrolyse der eigenen
Piiospborsiiureestor der Blutkorpercben auch bei Zinimertcnipe-
ratur erwarten. In einoin Yersueb, in dem eine Hamolyse durch
Zusatz von Saponin bewirkt worden war, erbielt icb auch eine
unrnittelbar cintrctcndc Hydrolyse. Sebon nacb 3 Stunden bei
17° war die Spaltung deutlicb, wahrend sicb in den bei 37° auf-
bewahrten nicbt mit Saponin versetzten Proben nacb 7 Stunden
in der Hegel keinc Veriindcrungen nacbweisen licss. In diesen
letztcren Proben war die Zunnbme der Permeabilitat der Bint-
korpennembran wiilirond dieser Zeit offenbar nocb nicbt so weit
vorgeschritten, dass cine voIlstSndige Kommunikation zwiseben
Ervtbrozvten und Plasma erreiebt wurde.

v •

Dic.se Yerbiiltnisse gelten also fiir die Pbospbormonoesterase.
Das Vcrhaltcn der Ppopbospbatase unterliegt einer Untersuebung
und wird in einer spatcren IMitteilung der Diskussion unterzogen
werden.

Die enzymatisebe Spaltung des Lecithins wird durch die Be-
stimmung des LipoidpJiospborgehalts vcrfolgt. Hierbei erhalt
man niclit etwa einen Ausdruck fiir die Wirkung der Lecitbinase
A, also die Dbcrfuhrung des Lecithins in Lysolecitbin, sondern
fiir die Wirkung der Lecitbinase B, die Abspaltung der beiden
Fettsauremolckulc. Gleicbzeitig folgte icb in gcAvissen Fallen der
Steigerung Alkohol-Ather-losbcher Fettsauren. Dicse braueben
jedoeb nicbt allein vom Lecithin herzustammen, sondern konnen
auch durch Zertcilung von Neutralfett und Cboiesterinestern
gebildet werden.

Der Lipoidphospborgehalt sank nacb 24 Stunden langer Er-
warmung bei 37° in einigen Versuchen im Plasma und auch in den
Erythrozyten. Im Zusammenbang mit der eintretenden Hamo-

130

KKUT SJOBEKG.

lyse wurde die SpaltungssctneUigkeit grosser und die Lecitkm-
menge sank kraftig in beiden Medien.

Die Mengen freier Alkobol-Ather-loslicber Fettsauren stieg
sowohl in den Erytbrozyten als auch im Plasma.

Demnach kommen, im Blut aucb Lecithinase B und die Dieste-
rase vor, die Cholinglyzerinphosnborsaureester spaltet.

Experimenteller Teil.

Die Versucbe wurden mit Pferdeblut ausgefiibrt. Die Bestim-
mung des Pbospbors gescbab nacb Fiske und Subbarow (1925)
mit Hilfe des pbotoelektriscben Kolorimeters.

Zur Bestimmung des Lecithins wurden die Blutproben nach
THEOBELii (1930) extrahiert und danach der Phosphorgehalt auf
die gewohnlicbe Art festgestellt.

Die iibrigen Blutbestandteile wurden nach den gebrauchlichen
Methoden analysiert. Die Analysen wurden sowohl mit Totalblut
als auch mit Plasma ausgefuhrt, und nach der Bestimmung der
Hamatokritwerte berechnete ich die Konzentrationen in den
Blutkorperchen.

Die Aufbewahrung der Blutproben geschah einen bis mehrere
Tage nach Zusatz von Na-Zitrat oder Heparin zur Verhinderung
der Kogulation teils bei 37 ° C und teils bei Zimmertemperatur
(17°).

Das Blutkorpervolumen veranderte sich wabrend 48 Stunden
nicht in denjenigen Blutproben, welche bei der niedrigeren Tem-
peratur aufbewahrt wurden. Bei langerer Aufbewahrung trat
Hamolyse ein, un3 dabei kam naturlich eine Verminderung der
Hamatokritwerte zustande.

Figur 1 zeigt die Zunahme der Hamatokritwerte in seeks Blut-
proben, die bei der hoheren Temperatur aufbewahrt wurden.
Der linke PfeUer in jeder Gruppe gibt den Wert unmittelbar nach
der Blutprobenentnahme an und der rechte nach 24 Stunden
langer Aufbewahrung der Probe bei 37°. In mehreren Fallen war
wabrend dieser Zeit bereits ein gewisser Grad von Hhmolyse einge-
treten, was naturlich eine Senkung des Hamatokritwertes herbei-
fiihrt. Bei weiterer Aufbewahrung stieg der Hamolysegrad
schnell, und nach 2 — 3 Tagen war die Hamolyse vollstandig.

Das Blutkorpervolumen nahm wabrend der ersten 24 Stunden
durchschnittbeh im Verhaltnis 1 ; 1.6 zu, also bis zu einem etwas
niedrigeren als dem berechneten Werte 1 : 1.75.

PUOSPHORVERBINDUNGEN UND PHOSPHATASE IM PFERDEBLUT. 131

Fig. 1. Hamatokritwerlo von 6 Pfcrdoblutprobnn. Linker Pfeiler vor der Er-
■vrfirmung. Rccbter Pfeiler nnch 24 Stundcn bci 37°.

Es lasst sicli denken, dass die Jr.sache der Diffusion der Fliis-
sigkeit in das Blutkorperehen liinein in einer Steigernng des os-
motischen Druckcs auf Grund eines Ereiwerdens von lonen liegt.
Es ist nicht moglich, den osmotischen Druck direkt im .Blutkor-
perchen festzustellen. Die Bestimmung des Gefricrpunktes des
Plasmas vor und nach der Erwarmung ergab, praktisch genom-
men, denselben Wert, vreshalb der osmotisclie Druck im Plasma
jcdenfalls keiner Veranderung anheimgefallen war.

Eine wahrscbeinlichere Erklarung der Diffusion ist eine Veran-
derung der Permeabilitat der Blutkorpermembran. Zur Unter-
suchung, ob cine solche Veranderung tatsacblich stattgefunden
liatte, wurde teils die Kesistenz der Blutkorperehen in Koch-
salzlosunger- verschiedener Konzentration, teils der K-, Na-,
Ca- und Cl-Gehalt im Plasma vor und nach der Erwarmung
bestimmt.

Vor der Behandlung fand sich der Beginn schwacher Hamolyse
in O.o5-proz. NaCl-Lbsung, und vollstandige Hamolyse trat in
0. 4-proz. Losung ein. Nach der Behandlung waren die entsprechen-
den Ziffern 0.75 und 0.6. Die Kesistenz war somit gesunken.

Die Kesultate der Analysen der obengenannten lonen finden
sich in Eigur 2. Wie vorher erwahnt, wurde in der Kegel Natrium-
zitrat zugesetzt. Dieses bewirkte eine Steigerung des Na-Gehalts
und einen Eingriff in das osmotische Gleichgewicht zwischen
Plasma und Erythrozyten. Trotzdem liess sich in den KontroU-
proben bei 17° keine Diffusion von Na-Ionen oder anderen lonen
feststellen. Eigur 2 zeigt jedoch, dass in den erwarmten Prober.

132 KNUT SJOBERQ.

Fig. 2. Die Pfeiler geben die verschiedenen Stoffe in rag im Plasma von 100 ml
Blut vor der Erwarmung und nach 24 Stunden langer Erwarmung auf 37° an.

die genannten lonen, durch die Blutkorpermcmbran wanderten,
Na •, Ca • • und Cl' in Richtung vom Plasma nach den Blutkorper-
chen, K • in entgegengesetzter Ricktung. Dieses hangt mit den
urspriinglicheu Konzentrationen dieser lonen zusammen, niir K '
kommt im Lbersclmss in den Erythrozyten vor. Die Yerminde-
rung des Gehalts an Na % Ca ’ ' und Cl' ist im Durcbscbnitt propor-
tional der diffundierenden Wassermenge, was also zeigt, dass die
Blutkorpermembran fur diese lonen vollig permeabel geworden
war.

Betreffs der Phosphationen stellt sich das Verhalten komplizier-
ter, da man bier nicht nur mit einer Diffusion zu tun hat, sondern
die Hydrolyse der Phosphorsaureester vorherrscht. Tabelle'l
zeigt einige Beispiele fiir die Veranderungen der verschiedenen
Pbosphorverbindungen. Est. P bezeichnet denjenigen Phosphor,
welcher als organischen Phosphorsaureester gebunden ist (Le-
cithin-P nicht mitgerechnet).

Bei 17° findet eine langsame Zerteilung der Phosphorsaureester
wahrend der beiden ersten Tage statt. Dieses beeinflusst die
Menge an anorganischem P im Plasma nur unbedeutend, da die
Permeabilitat fiir Phosphationen nur gering ist. Hach 2 — 3

PnOSPnOBVERBIKDDXGEN UKD PIIOSPIIATASE IM PFERDEBLUT. 133

Tabellc 1.

Die Vcrtcilung dcr Phospliorverbindungen itnd der Pliospbormonocstcrase zwisclien
dcm Plasma und den Erjdhrozj’teix im Pfcrdeblut unter vcrschiedenen Bedingungcn
in vitro. Die 'W'erte gebon die Jlcngon im Plasma und Erytlirozyten vom 100 ml

Blut in mg an.

Erythrozj'ten

p Phos-
phata-
S soB.E.

Bemerkungen

4..’iS 17°

1.31 Hamolyso

0.19 Kriiftigo Haniolyse

4.581 37°

2.08 Hamolyso
— Vollstiindige Hiiniolysc

6a 0| 36
24 37
48! 37

6b 0| 36
124 45
|48l 0

0 3
24 3
48

20 mg Saponin pr 100
ml. 17°. Vollstandige
Hiimolysc,

37°

Ein wenig Haraolyse
Kruftige Hiimolj’se

17°

Ein xvenig Haraolyse

37°

Ein u'cnig Haraolyse

Ein avenig Haraolyse 1

37°

ICriiftigc Hamolyso j
Vollstandige Hamolyse '

17°

Ein avenig Hamolyse |

37° i

Hamolyse !

Vollstandige Hamolyse '

37° ;

Hamolyse |

Vollstandige Hamolvse

134

KNOT SJOBBRG.

Tagen beginnt allmahlicli eine TLamolyae der Blutkorpercben
stattzufinden, und bierbei tritt der Ester-P ins Plasma iiber und
die Menge an anorganischem P nimmt in demselben zu. Mit der
Ausbreitung der Hamolyse auf den grosseren Teil der Blutkor-
perchen steigt die Hydrolysescbnelligkeit und erreicht in gewissen
Fallen relativ bohe Werte. (1 a, 2 a, 4 a).

Bei 37° ist die Hydrolysegescbwindigkeit schon innerbalb 24
Stunden hoch, aucb wenn eine offensichtlicbe Hamolyse noch
nicbt stattgefunden bat. Die Menge anorganiscb gebundenen Pbos-
pbors, die -wabrend der ersten 24 Stunden frei wurde, variierte
zwiscben 44 und 92 Prozent der urspriingbcben Quantitat. In
Versucb 1 b trat keine Zunahme der Pbospborverbindungen im
Plasma nach 7 Stunden ein, was beweist, dass sich die Permeabili-
tat erst andern muss. Nach 48 Stunden waren die Phosphorsau-
reester in einigen Fallen voUstandig bydrolysiert worden (4 b,
6 b). Scbon nach 24 Stunden war der Pbospborgehalt im Plasma
gestiegen, was darauf berubt, dass die Phospbationen aus den
Erytbrozyten je nach dem Grade diffundierten, in welch eni die
Konzentration zunabm.

In den Fallen 3 b, 5 b und 6 b war nach 48 Stunden langer
Erwarmung die totale Menge an saureloslichem P gestiegen, was
auf ein Freiwerden von Lipoidphospbor hindeutet.

Vcrsucb 2 c zeigt den Verlauf der Phospborsaurebydrolyse
nach der Hamolyse mittels Sapom'ns. In diesem Falle setzte die
Hydrolyse unmittelbar ein, obgleicb die Blutprobe bei 17° auf-
bewabrt wurde. In 24 Stunden waren ungefahr 23 Prozent der
Pbospborsaureester gespalten worden, im unbamolysierten Blut
dagegen nur 11.5 Prozent, was also einer halb so grossen Hydro-
lysescbnelligkeit entspricbt.

In einigen Fallen wurde aucb die Pbosphormonoesterasewirkung
bestimmt. Es ergab sich, dass die Pbospbatasewirkung in den
Erytbrozyten abnimmt, aber im Plasma im Zusammenbang mit
der Hamolyse ansteigt. Die totale Pbospbatasemenge andert
sich im Totalblut anfangs nur unbedeutend, sinkt aber bei larige-
rer Aufbewabrung. Dieses zeigt, dass die Pbospbatasewirkung,
die bei der Analyse zum Ausdruck kommt, nicbt der Enzymmenge
entspricbt, welche im Blut zur Wiikung gelangen kann. Hierbei
ist jedocb zu beacbten, dass gewisse Pbospborsaureester von dem
bier bestimmten Enzym nicbt gespalten werden. Das Pyro-
pbosphat wird von der Pjrrophospbatase bydrolysiert, aber dicse
Art von Estern bildet nur ungefahr 10 Prozent der totalen Ester-

PHOSl’nORVERBIKDUXGKN UKD PHOSPHATASE IM PFERDEIiLUT. 135

menge. Die Phosphormonoesterase im intakten Blut sclicint sick
deshalb in einem solchen Z\istande zu befinden, dass ilire Wirkniug
nicht- zu ihrem vollen Reclit kommt, aber niit der Zunahme der
Pcrmeabilitat der Blutkdrpermcmbran und mit eintretender Ha-
molyse kann sie ihren vollen Einfluss entvickeln.

Im Versuch mit Saponinkiimolyse erhielt ick bei der Bestim-
mung der Phosphatasewirlamg einen etwas hoheren Wert als bei
der Bcstimmang in unvcriindertem Blut. Der XJnterschied 0.9
B. E. ist jedock nickt so gross, dass cr die Versckiedenkeit in der
HydrolysesckuelHgkeit zu erklSrcn vcrniag, sondern durfte einem
Versuchsfekler zuzuschreibcn sein.

Tabclle 2.

Die Phosphomonoestcrasc in Bodnnsky-Einheifccn (B. E.).

Xr.

Xach

BODA^•SKY

In physiol.
Kochsnl/.los.

Bemcrkungen

5.44

•■>.14 i Blut

8a! •'5.52

8b! G.85

5.52

6.71

Blut

Blutkorp,

9rt| 10.20 1 —

9b! 8.01 1 —

Blutkorp,

Hamolysat.

lO.lj 9.08

10b} 10.80

lOcl 7.18

9.92

Blut

Blutkorp.

Hfvmolysat.

Tabelle 2 zeigt das Resultat der Bestiramungen der Phosphor-
monoesterasewirkung im Blut, toils auf gewoknliche Weise nack
Boda^'sky, teils nack derselben Methode. aber mit einem Zusatz
von 0.8 Proz. NaCl. Bei der ersteren Versuclismetkodik tritt eine
Hamolysc der Blutkdrperchen ein, im letzteren Fallc nickt. In
einigcn Fallen vurde die Phosphatasewirkung direkt in einer
Aufsckvemmung mittels pkysiologiscker Kocbsalzlosung ge-
waschener Erj’tkrozyten bestimmt. In alien diesen Fallen erkielt
ich innorkalb der Versuchsfekler dieselben Werte. Die in den Blut-
korpercken vorkandene Phosphatase kann also auf die Glyzerin-
pkospkorsaure der Reaktionsmisckung ikre voile Wirkung ent-
falten, okne dass die Blutkdrperchen kiimolysiert werden. Dieses
deutet darauf kin, dass sick die Phosphatase auf der Oberflacke
der Blutkdrperchen befindet und mit dem umgebenden Medium
in naherem Kontakt steht als mit dem Inhalt der Blutkdrperchen.
Die Versuche 9 b und 10 c zeigten einige Falle, in welchen die
Phosphatasewirkung bestimmt wurde, nachdem die Blutkdrper-

136

KNUT SJOBERG.

chen mit destilliertem Wasser bamolysiert und die Stromata ab-
filtriert wordeii waren. In diesen Vcrsuchen fand sicli alle Phos-
phatase im Hamolysat. Bei der Hamolyse -wird die Phosphatase
also aus der Blutkorpermembran freigemacht.

Da es gait, die Veranderungen, denen das Lecithin anheimfalit,
anal 3 '^tisch festzustellen, bestimmte ich die Verminderung im Li-
poidphosphorgehalt. Dieses bildet in ei'ster Linie einen Masstab
fur die Abspaltung der beiden Fettsauren aus dem Lecithiu-
molekiil, also fiir die AVirkung der Lecithinase B. Fine andauernde
Zerteilung unter Freiwerden von Phosphorsaure gemass den For-
meln auf Seite 126 ist schwieriger nachzuweisen, da man nicht
entscheiden kann, ob die freigewordene Phosphorsaure vom
Lecithin oder von anderen Phorsphorsaureestern herstammt.
In gewissen Fallen Hess sich jedocH eine Zunahme der Menge sog.
saurelosHchen Phosphors nach drei Tage langer Behandlung
nachweisen, und diese Steigerung muss auf vorher vorhandenem
Lecithinphosphor beruhen. Eine chemische Bestimmung von
Lysolecithin lasst sich in geringeren Blutmengen ebenfalls nicht
ausfiihren. Die Bildung von Lysolecithin ist mit der Veranderung
der Form der Erythrozyten und dem Eintritt der Hamolyse als
festgestellt zu betrachten.

Eine Abspaltung der Fettsauren kann ebenfalls analytisch
durch Titrierung von Alkoholatherextrakt nachge\viesen werden.
In einigen Fallen wurden solche Bestimmungen auch ausgefiihrt.
Die freigewordenen Fettsauren kbnnen jedoch nicht nur vom
Lecithin herstammen, sondern auch von Neutralfett und Choles-
terinestern, weshalb diese Bestimmung keinen Masstab allein
fiir die Zerteilung des Lecithins bilden kann.

Bei 17° Hess sich in Pferdeblut innerhalb 3 Tage keine Leci-
thinspaltung mit Sicherheit nachweisen. Figur 3 zeigt einige Bei-
spiele von Lecithinanalysen in Blut, das bei 38° aufbewahrt wurde.
Die Hydrolysegeschwindigkeit varuerte in verschiedenen Proben
und steht in gewissen Relation zu der Hamolysegeschwindigkeit.
Die Versuche 4 und 13 z. B. zeigten nur weuig Hamolyse, die A'^er-
suche 5, 6 und 11 hamolysierten dagegen schnell. Die Spaltung des
Lecithins findet sowohl im Plasma wie in den Erythrozyden statt.

Der Gehalt an alkoholatherloslichen Sauren stieg in Blutproben,
die auf Korpertemperatur erwarmt ^\'UTden. Die Zunahme fand
sowohl im Plasma als auch in den Blutkorperchen statt und nahm
mit dem Hamolysegrade zu (Tab. 3). Der pH-AVert des Plasmas
zeigte eine schwache Steigerung. '

PlIOSPnOEVERniNDUKGnN UKD PHOSPHATASE IM PFERDEBLUT. 137

22 Or

Nf. 3 i 5 b 7 n 12. O

Fig. 3. \'orrin(lrninpcn in tier rx'cithinmcngc. Dor gcfiillto Toil dor Pfcilcr gibt
die Menge in tlcn Hlutkilrpcrohcn nn, dor ungefillUe dicMcngo im Plasma von 100 ml
Hint vor der Ertv/irmung und tiach 24 bcr.w. 48 Stunden Inngcr Erwnrmung nnf 37®.

Tubello 8.

ml 0.1 n XnOH

Xr,

Stunden

pr 100 ml

Pii im

Blut

Plasma

of)

8.5

3.9

7.50

11.7

10.9

8.00

17.0

—

7.35

8.1

6.4

0.3

7.50

28.8

—

8.00

Zusararaenfassung.

Die Arbeit bcliandelt die Untersuclmng der Stabilitat organi-
scher Phosphorverbindungen und von Phosphatase im Pferdeblut
bei dessen Aufbewalirung in vitro bei 17° und 37° C wahrend eines
bis inelirerer Tage.

10 — ^i01323. Acta phys. Scandinav. VoJ. 1.

138

KNOT ejSBEBG.

1. Bei 17° fallen die Phospliorverbindungen im Pferdeblut
in 1 — 1 Tagen nur geringeren Veranderungen anbeim. Allmahlicb
tritt Hamolyse ein, nnd je nach dem Grade ihres Portschreitens
beginnen die organiscben Phosphorsaureester in den Erythrozyten
unter Abspaltung von anorganiscben Pbospbaten bydrolysiert
zu werden.

2. Bei 37° andert sicb die Form des Blutkorpercbens von der
miinzenabnlicben zur spbariscben Form, wodurcb die Hamato-
kritwerte in 24 Stunden durchschnittlicb im Verbaltnis 1 ; 1.6
steigen. Die Permeabilitat der Blutkorpermembran nimmt zu,
so dass K*, ISTa-, Ca* •, Cl’ und PO 4 ’ ’ ’ dieselbe ungebindert pas-
sieren konnen. Bei langerer Aufbewahrung bei 37° tritt Hamolyse
ein, die in der Regel nacb 3 Tagen vollstandig ist.

3. Schon 'wabrend der ersten 24 Stunden findet eine kraftige
Hydrolyse der organiscben Pbospborsaureester statt, und nacb
3 Tagen ist diese oft vollstandig.

4. Aucb die Lecitbine werden zerteilt, wobei teils die Fett-
sauren, teils die Pbospborsaure frei werden.

5. Pbospbormonoesterase kommt sowobl in den Erytbrozyten
als aucb im Plasma vor. Das in den Erytbrozyten vorbandene
Enzym beeinflusst eine Glyzerinpbospborsaure entbaltende Reak-
tionsmiscbung gleicb kraftig, sei es dass die Blutkorpercbcn bei
der Analyse bamolysiert sind oder nicbt. Dagegen scbeinen die
eigenen organiscben Pbospborsaureester der Erytbrozyten nicbt
angegriffen zu werden, bevor die Blutkorpercben gewissen Ver-
anderungen anbeimgefallen sind, die zu Hamolyse fiibren.

Literatur.

Beltanti, S., a. Contardi und A. Ercoli. Ergebn. Enzymforscb. 1936,
5 , 213.

Bergenhem, E., Acta path, mikrobiol. Scand. 1939. Suppl. 39.
Bobansky, a., J. biol. Chem. 1933, 101 , 93.

Fiske, C. H. and Y. Subbarow, Ebenda 1925, 66 , 375.

FAhraeus, R., Nord. Med. 1939, 1 , 885.

JosT, H., Z. Pbys. Chem. 1927, 165 , 171.

Kay, H. D., Biochem. J. 1928, 22 , 1446.

— , and R. Robison, Ebenda 1924, 18 , 1139.

Lohmann, K., Biochem. Z. 1928, 203 , 164.

— , Naturwissenschaften. 1929, 17 , 624.

Roche, J., Biochem. J. 1931, 25 , 1724.

Theorell, H., Biochem. Z. 1930, 223 , 1.

From the Women’s Clinic, University, Lund (Prof. A. Westm.-vn) and
the Women’s Clinic, Goncrnl IIo.spitn1, lifalmd (Dr. S. Genell),

Sweden.

Studies on tlic Muscular Pliysiology of
Die Genital Tract.

I. The Spontnneons Tonus of tlio Uterine Muscle;
Its Dependence on llormonic Factors.^

SDUE GENELL.

Our ido.ns concerning the importance of the sox hormones for
the tonu.s of the uterine mmscle arc cliiefly based on solitary ob-
servations that are not of a nature to admit of statistical treatment.
Certainly the. baseline of the curve obtained in the Magnus-
Kchrer experiment can give some idea of myonietrinl tonus, but
single observntion.s have very limited value, several widely variable
factors being involved, such ns the load applied, tlic composition
of the scrum saline solution and otl)ers. Most reports give no
particulars of these factor.s, and hence the observations made are
not comparable. There arc however solitary observations which
tend to show that ocstrin exercises a tonus-depressing action on
ut-crine muscle, at any rate in some animals, and it would therefore
appear desirable to determine whether this is the case. The pres-
ent investigation has been carried out on rats.

Method: A uterine cornu is extirpated under ether narcosis
and, wnthout being cut up, is disposed as a longitudinal preparation
by the routine method applicable to surviving organs. Serum
saline solution is used of the following composition: NaCl 8, KCJI
0.42, CaCl, 0.24, MgCl, 0.005, NaHCOs 1, glucose 0.5 per 1000 c.c.
Through tin's fluid is bubbled oxypn with 5 % COj. The kyrno-
graph is turned by hand at certain intervals. After the preparation
has been mounted, it performs its initial contraction, then relaxes,
and, when the recording pen begins to rise again at the second
contraction, the kymograph is manually turned about 1 centi-

’ Ecceived 17 Jnly 1940.

140

SUNE 6ENELL.

metre. Five minutes later the second turn is made, after which
the kymograph is turned every fifth minute for half an hour and
thereupon every tenth minute for 2 V 2 hours. The “curve” obtained
in this way consists of vertical curved lines, the lower points of
which denote the maximum relaxation, the upper points the maxi-
mum contraction, during the respective intervals of time. The
first curve (denoted “0 min.”) represents only the initial contrac-
tion and the relaxation following upon it. The initial contrac-
tion was the highest of all recorded during the experiments,
except in two cases.

A horizontal line is drawn through the top point of the initial
contraction curve, and the distances vertically from this line to
the relaxation maxima (i. e. the lower points of the curved lines)
during the respective intervals of time are measured and tabulated.
^^Tien the pen has a stroke ratio of 1:1, the true values are
obtained. The mean of the relaxation values within a certain series
of experiments is calculated for each time interval. A graphical
record of these values will give a curve that represents the degree
of relaxation of the cornu uteri during the three hours the experi-
ment has lasted. Tonus in- unstriated muscle is defined by Evans
as “a resistance of its substance to extension”. Consequently,
the relaxation will be less the higher the tonus is. The relaxation
curve described above will therefore give a picture of the. state
of tonus of the longitudinal muscle of a surviving cornu uteri.

All the factors involved in the experiments were standardized
as far as possible, the time selected for the experiments, operative
technique, technique aud time taken for the preparation, etc.
Identically the same instrument, the same recording pen and the
same load, balanced so that the pen just had overweight, were
employed. The length of time required for the investigation ne-
cessitated the use of two kymographs and hence two pens. The
very slight difference between the two instruments consisted in
the load being slightly larger in one of them. One half of each
series of experiments were made with the heavier loaded instru-
ment (group A), one half with the lighter loaded (group B). Hence,
although there is a distinct difference between groups A and B
(the figures are published in Genbll, 1937), no allowance need be
made for this difference on comparing the different means.

Each animal was used twice. The two uterine cornua were taken
in different sexual phases, but both were examined in the same
apparatus. This gave rise to still another grouping of the material:

Tnlilo 1

THE MUSCULAR PHYSIOLOGY OP THE GENITAL TRACT. 141

T =

180

in.

p p ws o
ci-r-l^cd

Cl Cl Cl Cl

p p p o

to to Cl

Cl »-«

O O vs o
o iri 00 ci

CO Cl Vi

p O o o

cd wi>^cd

v-t vi^AJI

p o p o
oedtded

Cl Cl Cl Cl

33.5

50.5

•TJ'

T =

120

tn.

o o o o
<£5^?i?CO

Gl Cl

WS p p o
odcdtdi^
Cl Cl Cl Cl

kS O p o
-•tdoo

Cl^»-^*-<

P »A US O
cicdtN'd'

Cl Cl -H

kA O kA kA

ci t>: td

•-I Vi Vi Cl

O kA kA kA
viwtdci
Cl CM Cl Cl

vi tv

cdCa

ClCl

T =

*^3 o wa

»ac>OCS
Cl VH Cl

p o p p

to ci CO

Cl Cl Cl Cl

kO O kA O

ci'ksicitd

Cl*-«Clk-t

p kA US O

flowed k-<

Cl Cl Vi

O O .A kS

kdwkdtd

Vi Vi V, Cl

O o kS kA

c>td\d<d

Cl Cl Cl -i

O US

Cl «i

T = 0 jn.

o o o o
cdci'r''^
^ <— ♦

kO O kS kA

o-i<cicd

k— t

O O kQ kS

S€d<^cd

Cl •-<

«A kA kA kA
TV* Ci O* O
•H Vi *-i

kA kA kA O

ci id vd cd

Vi Vi

e»

11.8

lO.I

*«

‘5

ocnKO

r- 1 '- i-»

iri CO •?* to
to to to to

r^*»?<csci

kfiotot^

t>-oor-kr*

W »f5 iCi rfi
tOdkCSCS

woo w w

CO to CO to
w o »rs o
Woo ww

coca -HO
WOOOOd
wwwco

■

T =

180 ».

•o o *n p
— Iwca ^
Cl ^ Cl

O O kO vs

ci c o5 *-<

Cl Cl Cl

p P o o
I*- iJl *-< to

p kA kA O

td •?»’ ci -kd

•-•Cl Vi

O tA A kA

o' ci cd

d Cl SI Cl

kA o kA kA

ededtdei

Vi Vi vi Vi

18.8

16.6

■

T =

120

■O kQ o

ci ci o

^Cl >»-• Cl

o o o o
ci

Cl Cl ^ Cl

p p O kA

cd "C? kd

p kA kA kA

W Cl* r-?

vi Cl Vi Vi

kA kA O kA

cd o ci '■ji
dCldCI

kA O O O

cd Cacdrfi

vi Vi Vi vi

00 CO

caw

vi vi

T = 60

in.

p o «o «.a

'-ior^cd

«-( Cl ^

O WS O vs
CCOCDC$
Cl Cl Cl

p kA kS kS

Cl id »d cd

Vi Vi

p p p p

edededkd

Vi (SI rH t-H

kA P O <0

ei o — «

Cl Cl Cl Cl

kO O kA O

vi cawed

Cl vi vi vi

cs »A
ow

Cl -H

-d

T =

= 0

O O O

p p O kA

td-j*— Jod

f-H •-< Vi

p kA O p

ci w w w

^ vi

p p p p

to coded

Vi Vi Vi Cl

US US O O

cd id o cd

vi V* vi

V- O

kd ci

vi vi

Animal

f'-Ote.-H

CfJOCOiTS

I'- r- r-

cccir^O

CliO’k^'tO

ccoOrK—<
tocoi^ ^
Wt^f^CD

«0 o Cl ifS
woooo

W OQ CO

CO to CO to
W O lO lO

woo w w

coca-iO

WCOCOd

wwwco

T =

180

\d ^ o

ci*^-^ ci

Cl « Cl Cl

O O kO vS

O O kA O
to COW
CS Cl Cl Cl

O CS kA kA

edo-Mid
Cl CO Cl Ci

O O O kA

•^5 wedo

CO Cl 51 Cl

kA O O kA

fld td id ci

Cl 51 Cl Cl

krf

tdkd

CiCl

T=:

120

vs kA p
ClCC^Cl
Cl r5 Cl Cl

kO VS o o

Ca to

Cl Cl Cl-<

O O kA k9

to w wtd
CC Cl Cl Cl

O kA kA kA

tdei vi-j5
Cl Cl Cl Cl

O kA kA kA

cd td cd ci
CO Cl Cl d

US kA US us

Wkdedei
d Cl d Cl

»d

p p

tdkd

ClCl

•<-

tr.

T = 60

TO.

i3 p p O

01 CC Cl Cl

kS »o »s o
caccrttd
Cl ClCl^

kA kA kA p
'•Jitdtk^td
CCClClCl

kA <0 kA O

idadcied

Cl Cl -4 Cl

o yi o o
(fi ci

codcici

O kA O kA
widfdvi
d CM Cl Cl

»d

cc es

kfjii?

dci

T = 0

tC p O *3

Socici

•- « Cl

o O kO ws
toed to z$

Cl « *-<

kA O kA O

c> *d Cl ci

CC Cl Cl

O O kA kA

«>©c>cd

Vi Cl Cl «

o kA kA kA

ci'0'’^’td

Cl Cl — Vi

O kA kA O

ci -i td o*

Cl Cl *H vi

Cl-vi

Cti

’S

cn«ci-i

rfO<fiC£i

••kC-ikTitO

flocir>*oo

l- t>»

SSliS

w w wee

to o Cl to
woooo
woo WO)

w^jdci
\0 o to w
wco ww

WkJ^VO-^
todiod
WCO w w

T =

180

TO.

o o vs va
cDcdci-J

CQ-C'Cl ^

o O kS ko
ookdt^
•^ccccci

O »A »0 kA

kd o ci cd

CC-rCACl

O kA O O

cd cd Cl w

Cl Vi CO*-*

o p p P
vi W'^

CO Cl d Cl

p P p

tdkdkdo

ClCl»HCl

to Ck

cikd

COd

'*1'

T =

120

TO.

>a VS o

cC-?< Cl-r

*S kS vs o
cacscc^
cO Cl cc Cl

kA CA p O

^cicicd

CO CC CO Cl

kA kA tA O
w Cl vl to*
Cl viCOk-4

kA p p p

otd'^d'^

CO Cl d Cl

p P p p

td'^'^o

Cl Cl vi Cl

O Cl

d id
cc Cl

T = 60

TO.

o p O kO

cC -rci cC

o o o p
odesedr^

CC Cl CC Cl

•A kA tA p

cd w-^w
ccccccd

kO kO O kO

to Vi .-iiri
Cl-vCAvi

<c P P P
Citovicd
CO Cl Cl d

lA kA kA kA

idcdcdca

dClvirH

W p

T = 0

TO.

o »o O vs
ci — J to

ClCtk-kCl

p o p p
« ^ ^ Cl

o O P p
•-^kdtdcs

ClCSCl kH

O kA O kA

ci WWW

Cl Cl

kA kA O p

kd ci

Cl Cl Vi Vi

p p p p
edOvica
Cl Cl vi

Ci rs
cica

d vi

'iS

t'- 1'- 1—

coot^o

ClO'*rtO

ClCCCl
•vototo
wco ww

•itc^YrjtD

cocs woo
l>. tv t» W

coo wco
kO wcoco
w w w w

Its CO to
to to CO to
WWW w

■whole mat.

»-i Hi

oa oa

*H *H

O O

BQnOJt

cr*

142

SDNE 6BNBLL.

group I, in wHcii the cornu was taken first, group II, in which ifc
was taken last. The difference between these two groups lies in
the possibility of the second cornu being affected by the first
laparotomy. In fact, the figures (Table 1) show a distinct differ-
ence, those of group II being uniforml)’- lower than those of
group I.

Yariations in Myometrial Tonns during the Sex Cycle.

The number of animals employed was 48. Fig. 1. gives the
curves for the four sexual phases [denoted pro (24), pos (24), met
(24), neg 4 (24). Each value on the ciuwe represents the mean of
24 determinations.

An extract from the data upon which the curves are based is
given in Table 1, in which the relaxation values obtained from
each of the 96 experiments are furnished for four different points
of time, viz. 0, 60, 120, and 180 minutes.

These values must be regarded as highly interdependent, which
implies that the differences between the means of the different
series at the different times will be mainly the same.

The square of the dispersion in each series, calculated by the
formula

72 =— ^i;(x — m)S
n — i

is given in Table 1 (Square of the Standard Deviation).

The values of y vary from series to series, though scarcely
more than may be considered as due to pure chance. The mean of
■all the values is 37, and the dispersion or standard error of the
single observation will therefore be 6.1.

The means of the values obtained for the different sexual phases
are given in Table 2. The standard delation of the differences

Table 2.

T= 60

T = 120

T= 180

Pro

21.1

27.6

28.6

29.3

Pos

20.2

25.0

25.7

25.9

Met

13.6

18.7

18.8

17.7

Neg 4

19.9

20.2

21.9

22.1

THE MUSCULAR rUYSIOLOGY OF THE GENITAL TRACT. 143

between different sexual phases is 1.7. Thus, there is no significant
difference betAveen pro-oestrus and oestrus or between met-oestrus
and dioestrus. The other differences are significant,

Myoiuetrial Toiins in Castrated Animals and in Oestrin-

treated Castrates.

The number of animals employed was 24. The castration was
always performed in the met-oestrous phase. On the day the first
cornu was taken into the experiment, the animal was administered
a certain amount of oestrin immediately after the operation. A
certain time after that, the second cornu was taken into the experi-
ment. The material includes two series, described as Old Series
and Neio Scries, which are not perfectly alike.

Old Series. Cornu I was extirpated a varying number of daj'S
(6th to 12th) after the castration. After the operation the animals
were given intramuscularly 5000 I.U. of oestrin in 0.1 cc oil so-
lution. Cornu 11 was then taken on the day the vagina] smear
presented a pro-oestrous picture, about 48 hours after the hormone
injection.

Ncio Series. Cornu I was alwaj^ taken on the 6th day after the
castration. After the operation the animals were given subcutane-
ously 300 I.U. of oestrin in an aqueous solution of alcohol. Cornu
II was taken 24 hours later, when the vaginal smear exhibited a
commencing pro-oestrous picture and the uterine cornu appeared
distinctly vitreous.

The data selected for statistical treatment are submitted in
Table 3, in which the relaxation values obtained from each of
the 48 experiments are given for four points of time, 0, 60, 120
and 180 minutes, together with the means of these values. The
table also contains the means of the values divided between the
two somewhat different series, the old series and the new series,
as well as the calculated square of the standard delation. The
standard eiTor of the single observation is 4.4. The standard errors
of the means for a mean of 24 and 12 individuals respectively will
therefore be 0.9 and 1.3 respectively.

Between the old series and the new series there is a difference
as regards the means for the imtreated castrates in the group.
It is however not a significant one and manifestly depends upon
chance.

A comparison of the fall in relaxation values from the series of

144

SUNK GENELL.

Table 3,

Untreated castrates

Treated castrates

T= 120

T= 180

T= 0

T= 180

802

20.5

21.5

21.6

25.6

32.0

32.5

33.0

812

11.6

12.0

13.0

12.5

19.0

20.6

21.0

831

lO.O

18.5

18.0

23.5

28.0

28.5

29.6

843

16.5

21.0

22.0

20.0

17.5

22.5

23.0

23.5

845

14.5

21.0

23.0

20.5

24.0

28.0

28.6

851

9.0

16.0

14.0

28.0

29.0

29.6

785

7.0

16.0

15.5

12.5

18.0

17.5

807

6.5

16.5

14.6

12.6

11.6

25.0

27.0

823

9.6

16.0

17.5

18.0

12.6

21.6

23.6

25.0

839

13.0

21.6

21.0

19.0

7.0

23.0

23.6

25.0

840

11.0

19.0

19.5

19.0

26.0

848

14.0

18.5

19.0

16.5

13.5

19.0

20.0

846

17.0

23.5

24.6

24.0

21.0

24.0

24.5

14.5

23.0

24.0

18.0

20.6

21.0

20.5

876

lO.O

14.5

16.5

17.5

31.0

34.0

34.6

35.6

864

14.0

23.5

25.0

25.6

21.0

26.5

27.0

27.6

861

16.0

19.0

20.0

21.0

22.0

24.0

24.5

880

13.0

22.5

23.5

24.5

31.0

32.0

33.0

863

8.0

17.0

21.6

22.0

21.6

22.5

22.6

856

19.0

26.0

26-5

24.0

27.0

17.0

17.6

15.0

859

lO.O

21.5

21.0

18.6

20.6

25.6

865

20.0

24.0

24.6

25.6

13.6

21.5

869

12.0

16.0

17.0

2C.0

24.0

24.6

26.0

881

16.5

26.0

27.0

27.6

23.5

24.0

24.5

24.0

whole mat.

12.4

19.7

20.5

19.9

24.4

25.0

25.2

a .

“S 1

Old series .

10.6

18.0

18.3

17.3

24.3

25.0

25.6

New series .

14.2

21.4

22.6

22.5

24.6

25.0

Square of Stand,
dev

17.2

14.2

15.7

17.1

40.4

20.0

19.7

23.6

. untreated castrates to the series of treated castrates gives the fol-
lowing mean values:

Old Series 7.0 ±

New Series 4.1 + 1.8

TUE MUSCULAR PHYSIOLOGY OP THE GENITAL TRACT. 145

As -will be seeu, this difference may be due to pure chance. The old
and ne-w series can therefore be lumped together without disad-
vantage.

The values obtained for group II (cornu II) are, on an average,
2.9 below those for group I (cornu I). With regard to the non-cas-
trated animals in different sexual phases (Table 1), the series con-
sist half of I-determinations, half of Il-determinations. As regards
the castrates, however, the series consist of either only I-deter-
minations (the untreated) or only Il-determinations (the treated).
For interserial comparison, therefore, the values for the untreated
castrates (which all belong to group I) must be reduced by 1.5
(strictly one-half of 2.9), and the values for the treated castrates
(which all belong to group II) increased by 1.5,

Table 4.

T = 60

T = 120

T = 180

Untreated

10.9

18.2

19.0

18.4

Treated

21.4

25.9

26.5

26.7

Table 4 contains the values lumped together and corrected in
the mant’or described. The difference between the untreated and
treated series is significant. The whole material, corrected as in-
dicated above, is grapliically represented in the curves “Untr.
Castrates (24)” and “Tr. Castrates (24)” in Fig. 1.

Discussio 7 i. E. Kehrer appears to be the first to have pointed
out the occurrence of tonus changes in the uterus. He adopts the
terminological conceptions derived from special investigations on
the phenomenon of tonus, and after him the term “tonus” often
occurs in the literature dealing with myometrial motility. Almost
equally often no definition or description is given of what the
author in question means by the term. In reports on the registra-
tion of the motility of the surviving organ, references are made to
increased or diminished tonus, usually as a result of drug action
in one or the other direction. To distinguish during observation
of the organ in situ between form-changes of a tonic nature and
such caused by active contractions is manifestly impossible. As
a consequence there has been some confusion of terms in the
literature, alterations in the state of contraction ha^dng been

0 60 120 teo min.

Fig. 1. State of tonns in the longitndinal mnscle of the nterns dnring 180 minutes.
The curves have reference to the four sexual phases as well as to untreated and
oestrin-treated castrates, 24 animals in each group.

confounded witli changes in the state of tonus and the two states
having sometimes been regarded as identical.

Starting from a given definition of the term “tonus” (p. 140),
tbe present investigation is an attempt to gather data suitable for
statistical treatment and indicative of the spontaneous tonus of
the uterine muscle and its dependence on hormonic factors.

The curves in Fig. 1 give tbe relaxation values for the longitu-
dinal muscle of the surviving cornu uteri between 0 and 180 min-
utes for the four sexual phases and for untreated and oestrin-treated
castrates. The degree of relaxation being dependent on the re-
sistance of the muscle to extension (here = the experimental
load), a curve should mirror the state of tonus of the longitudinal
muscle during the stated space of time for the respective phase,
condition after castration, or condition after oestrin treatment of
castrates. If Evans’ previously quoted definition of tonus is ac-
cepted, the curves may be denoted as “Tonus Curves”.

As was made clear in the statistical discussion above, there is
no significant difference between the tonus curves for pro-oestrus
and oestrus, nor between those for met-oestrus and dioestrus.
Since the pro-oestrous and oestrous phases coincide with the
maturation and rupture of the follicles, the incretion of oestrin
must be highest during these stages of the sex cycle. In the met-
oestrous and dioestrous phases the corpora lutea ovulationis have

THE MUSCULAR PHYSIOLOGV OP THE GENITAL TRACT.

147

Fig. 2. VariaHons in the loniis of the longitudinal mnsdo of the nterns during
a eexnal rycle. (.SLago I — V = acxnal phnacs according to Long and Evans’, 1923.
Pro-Kcg 4 sexual phases .according to the author’s own division of the sexual
cvcic. Pro = pro-oeslniB, pos = oestrus, met = met-oestrns. neg 4 — dioc.strns.)
The figure.s arc the means of the values found for 60, 120, and 180 minutes.

been formed, New follicles do not begin to grotv until towards the
end of dioestrus. The oestrin incretion is therefore lowest in these
stages. For the purpose of this discussion it is therefore practical
to place pro-oestrus and oestrus together under the common head
of '"Positive Phase" and, similarly, met-oestrus and dioestrus under
the head of "Negative Phase". This grouping is biologically mo-
tivated by the ocstrin-incretnry conditions mentioned above, and
al.so finds .support in the fact that tbe tonus curves for the respec-
tive pairs of sexual phases do not exhibit a significant difference,
although, of course, this does not rule out the possibility of such a
difference being found on the basis of a larger material.

On the other hand, there is a significant difference between
one or the other of the tonus curves for the positive phase and one
or the other of the tonus curves for the negative phase. The pos-
itiv'c phase or heat qiliase is thus characterized by a substantially
lower myomclrial tonus than the negative phase.

Fig. 2 presents a graphical view of the tonus variations during
a sexual cycle. The relaxation values for the respective sexual
phases have been computed as the values for 60, 120 and 180
minutes. The value for 0 minute has been omitted for the reason

148

SUNE GENELL.

that at this juncture the organ, cannot yet he considered to have
adjusted itself to the experimental conditions, a fact that is clearly
shown by the curves in Fig. 1.

In Fig. 1 it is observable that in the main the tonus curve for
untreated castrates coincides with the tonus curves for the negative
phase, and the tonus curve for oestrin-treated castrates in the main
with the tonus curves for the positive phase. It is thus possible,
by treating castrated animals with oestrin, to produce experi-
mentally the low state of myometrial tonus characteristic of the
heat phase.

Summary.

Myometrial tonus varies with the sexual phase. It is low'
in the heat phase, high in the quiescent phase. The faU of tonus
in the heat phase is elicited by the heat hormone of the ovary,
oestrin.

I\Tyometrial tonus is not influenced by castration as such. It
remains unchanged within that space of time following castration
during which muscular atrophy has not yet appeared.

It is probable that the uterine muscle possesses a certain state
of normal tonus, maintained largely by unknown factors. We only
know that the tonus of the uterus depends on ionic factors, on the
osmotic pressure of the blood, and on the hydrogen ion concentra-
tion, but little is known of the mechanism through which these
factors act. Under the action of oestrin the normal state of tonus
falls during the he;it phase. The mechanism of this oestrin action
is unknown.

Eeferences.

Evaxs, C. L., Physiol. Rev. 1926, 6, 358.

Genell, S., Uterin- och vaginalmuskulaturens funktionella uppgifter
i den icke-gravida organismen. Lund, 1937.

Keheer, E., Miinch. med. Wschr. 1912, 1831.

Long, J. A., and H. M. Evans, Mem. Univ. California 1922, 6.

From the Carlsberg Laboratory, Copenhagen.

Method for Rapid Determination of Specific

Grayity.i

O. P. JACOBSEN and K. LBSTDERSTRCM-LANG.

(With 1 fig. in the text.)

The method here described is a simplification of that developed
by Linderstrom-Lang and Lanz (1938).

Principle.

A reasonably linear specific gravity gradient is produced in a
vertical measuring cylinder (200 cc) by mixing kerosene and
bromobenzene in varying proportions. If a drop of a given so-
lution is introduced under the surface, it will fall \7ith dimin-
ishing velocity and finally come to rest at a position in the cyl-
inder where the specific gravity of the kerosene-bromobenzene
mixture is equal to that of the drop in question.

Before introducing the drop of the solution under investigation,
other drops (standard drops) are introduced of potassium chloride
solutions of accurately known specific gravity. Naturally these
drops will also come to rest at levels where their specific gravities
are equal to that of the surrounding medium. Plotting the posi-
tions of the potassium chloride drops as ordinates against the cor-
responding specific gravities as abscissae, a reasonably straight
line is obtained in the coordinate system, from which it is possible,
knowing the position of the drop of the unknown solution, to read
the specific gravity of this solution with considerable accuracy.
The specific gravity may also be calculated by linear interpolation.

This simple principle is known from the above quoted paper
by Lindebstrom-Lang and Lanz.

’ Received 14 July 1940.

150

C. P. JACOBSEN AND K. LTNDERSTR0M;-LANG.

The additional facts which justify a separate publication are
that for rough density determinations it is unnecessary to place
the gradient tube in a thermostat and that the reading of the
positions of the drops may be done with the naked eye making
use of the scale divisions oh the measuring cylinder which serves
as gradient tube.

Preparing the (xradient.

If it is the object to determine specific gravities of, say, from
1.03 to 1.11, two kerosene-bromobenzene mixtures are prepared,
of the specific gravities abt. 1.01 and 1.12 respectively. 100 cc

1.00 1.02 1.04 1.06 1.08 1.10 1.12

densities

RAPID DETERMINATION OP SPECIFIC GRAVITY. 151

of tlie heaviest mixture is filled into the measuring cylinder,
whereupon 100 cc of the lighter mixture is cautiously filled in
on top of the heavy one. In order to adjust the gradient, a long
spatula is moved up and down through the cylinder and rotated
at the same time, until streaks appear at the top and at the
bottom. After standing for a short time (30 min.) the gradient
tube is ready for use.

Upon standing, the gradient curve becomes flatter because
the differences in specific gravity decrease due to currents and
diffusion. The end result will be that the specific gravity is the
same at all heights. This, however, requires a rather long time.
The figure shows the change of the gradient in the course of 14
days for a gradient tube which stood unprotected on a table,
exposed to vibrations and drafts.

Making the Determinations.

The standard drops, having specific gravities varying from
1.03 to 1.11 with intervals of 0.02 are introduced by means of a
1 mm® pipette with rubber tube. (Compare Linderstrom-Lang
and Lanz). In this rough method the drop size is of little import-
ance and may vary 100 per cent without affecting the equilibrium
position. After the pipette has been filled with the standard
solution and emptied again by blowing, (not of course, into the
bottle of the standard solution) a number of three times, it is
filled to the mark and dried on the outside with filter paper,
without the paper touching the point. The drop is then intro-
duced into the gradient tube by the following method: The point
of the pipette is brought to abt. 3 mm below the surface; the
drop is forced out by blowing gently through the -rubber tube,
but does not slip from the point until the pipette is cautiously
•tvithdrawn through the surface, while at the same time care is
taken to have a low excess pressure in the pipette. It is then
possible to introduce a second standard drop, of the same speci-
fic gravity, without making the above mentioned three cleaning
operations. The next standard is then introduced, after three
cleanings. It is obvious that the heaviest standard drops should
be introduced first in order to make collisions between the drops
less likely.

Drops from the solution to be examined are then introduced
in the same marmer.

152

C. F. JACOBSEN AND K. LTNDERSTE0M-LANG.

Eeading of the height of a drop is made by keeping the eye at
the level of the drop, in order to eliminate errors due to paral-
laxis.

A number of drops may, of course, be introduced into the gra-
dient tube at the same time, but it is not advisable to use the
tube for more than abt. two hours at one time.

All drops are removed when there is no more room for drops
to be tested or when abt. two hours have lapsed since the standard
drops were introduced. For this purpose a long, thin glass rod,
with its point wrapped in moist filter paper, is introduced into
the gradient tube; the drops are absorbed by the filter paper as
soon as it touches them. The gradient tube is again ready for
use after standing for abt. 20 minutes.

The method may, for example, be used in determining the
specific gravity of blood. Its accuracy is 0.1 per cent.

Reference.

Linderstb 0 M-Lang, K., and H. Lanz, C. B. Lab. Carlsberg, ser. chim.
1938, 21. No 24,

From the Pharmacological Department, University of Lund.

On tlie Influence of Atropine on some Nicotine-
like Actions of Acetylcholine.^

N.-O. ABDON.

(^Vitli 4 fignrcs in the test.)

In his fundamental study on the pharmacology of acetylcholine
(ac. ch.) Dale (1914) divides its effects in muscarinelike and
nicotinelike actions. The susceptibility to atropine was considered
a main difference between the two kinds of action. The muscarine-
like actions were abolished by atropine, while the nicotinelike
ones were supposed to bo uninfluenced. Since then, the nomencla-
ture and classification of Dale have been widely accepted, but
with regard to the atropine antagonism the classical definition
has been modified dc facto. Thus, Riesser and Neuschloss (1921)
showed that atropine abolished the effects of ac. ch. on skeletal
muscles of frog, and in 1930 Dale and Gaddum found that atropine
prevented or diminished the action of ac. ch. on stripes of cat’s
denorvated diaphragm, suspended in Ringer’s solution. Frank
Nothmann and Hirsch-Kauffmann (1922) found that scopola-
mine antagonized contractures of skeletal muscles provoked by
intraarterial injections of ac. ch. This result could not be confirmed
by Dale and Gaddum (1930).

As a general rule, both kinds of ac. ch. actions are antagonized
by atropine, although several exceptions are described. Thus,
the action of ac. ch. on the heart of helix pomatia (Jullien and
Morin, 1931) and on the dorsal muscle of leech (Minz, 1932) are
not influenced. Although atropine antagonizes the ac. ch. effects
on striped muscles of frog and rabbit when suspended, it has
hitherto not been possible to demonstrate the antagonism on the
"quick contractions” of voluntary muscles, provoked by intra-

* Received for publication 4 Aug. 1940.

11 — i01823. Acta phys. Scandinav. Vol.J.

154

N.-O. ABDON.

arterial injections of ac. ch., neither in mammals (Browi^. Dale,
and Feldberg, 1936) nor in frog (Eaventos, 1937). In the
present paper some experiments are reported in which it has
been possible to abolish the “quick contractions” with reasonable
amounts of atropine. According to Bacq and Brown (1937)
atropine does not antagonize the action of ac. ch. on ganglia.

Although the nicotinelike as well as the muscarinelike actions
are antagonized, there is generally a quantitative difference be-
tween the t wo kinds of actions with regard to the amounts of atro-
pine necessary. Most muscarinelike actions are abolished by
small amounts of atropine, while most nicotinelike ones require
so large amounts that the antagonism has to he studied on iso-
lated organs. Many authors are of opinion that this quantitative
difference is so great that the action of atropine has still to be
accepted as a criterion of a fundamental distinction between
muscarinelike and nicotinelike actions. If one accepts this quanti-
tative difference as proving such a distinction, it would, however,
he necessary to disregard the striking regularity in the influence
of other agentia on the actions of ac. ch.

Whether an action of ac. ch. is nicotinelike or muscarinelike,
whether it is regarded as inhibiting or stimulating, it is potentiated
by drugs as eserine, fluoride, ergotamine, oxalate, citrate, quinine,
digitalis glucosides, etc. It is potentiated by degeneration of
efferent nerves (degeneration of motor nerve of striped muscle.
Dale and Gaddum (1930) et al.; of motor nerve of smooth muscle,
Rosenblueth (1932); and of preganglionic nerve of a ganglion.
Cannon and Rosenblueth (1936). Also hydrogen ions poten-
tiate the effects of ac. ch. (Andrus, 1924, et al.). Curarine, on the
other hand, antagonizes the action of ac. ch. on striped muscles
as well as on the frog’s heart (Raventos, 1937).

As to eserine and prostigmine, this regularity is explained by
the inhibition of the choline esterase, but the effect of the other
agentia cannot at all, or only to a small extent, he explained as
an influence on the esterase. Fluoride, ergotamine, and quinine
cause in rather large concentrations a certain inhibition of the
esterase, but this action seems to he of little importance to their
potentiating effect. Kahlson and Uvnas (1938) showed that these
drugs develop their potentiating effect in concentrations which
are too small to exert any inhibiting action on the esterase. They
found further that these drugs sensitize the organs to the stable
choline compound carhaminoylcholine. In this institute we have

INFLUENCE OP ATROPINE.

155

furtlier examined oxalate, citrate, and strophantin. We found
that a concentration of 1 : 1,000 of these drugs does not inhibit
the esterase to any measurable degree, while much weaker solu-
tions sensitize the eserinized m. rectus abdominis and the frog’s
heart to ac. ch. (not published). Nor can the sensitizing effect
of hydrogen ions be e.xplained as an influence on the esterase.
Ahlgren (1929) found that preparations of the rabbit’s gi>t is
markedly sensitized to ac. ch. by moving the pH of the suspension
fluid from 7.7 to 7.1, while the results of Wahlqist (1935) show
that variations in xhe hydrogen ion concentration in the range
between pH 8 and pH 7 have no effect on the esterase activity.
Nor is the effect of degeneration of efferent nerves caused by
decrease of the esterase acthdty. According to D. Naciim.\nsohn
the esterase activity of denervated sceletal muscles is increased
(not published). It seems, therefore, to be necessary to assume
that the majority of the potentiating agents exert their actions
by interfering with the mechanism of action of ac. ch.

The regularity in the influence of many agents on the different
actions of ac. ch. suggests an essential similarity between the
mechanisms of action of muscarinelike and nicotinelike effects.
Against this similarity stands the quantitative difference between
the various actions of ac. ch. in their relation to atropine. It
seems, therefore, to be of a certain interest to study if this quan-
titative difference in reality signifies a qualitative distinction
between muscarinelike and nicotinelike effects. One must,
namely, be impressed by the fact that the producing of nicotine-
like effects generally requires much larger concentrations of ac.
ch. than the producing of muscarinelike effects, and it seems to
be possible that those properties of an organ which make a large
amount of ac. ch. necessary also make a large amount of atropine
necessary for abolishing the ac. ch. effect. Clark, Gaddum,
el al. {vide Clark, 1937) have produced mathematical expressions
of the quantitative relations of ac. ch.-action and of the anta-
gonism between ac. ch. and atropine. On the basis of those ex-
pressions the writer has studied the ratio between ac. ch. and
atropine in a series of experiments on various organs of frog.

As described by Clark (1937), the relation between concentra-
tion and effect of ac. ch. follows Hitchcock-Langmuir’s law in
all organs studied:

156

N.-O. ABDON.

K is a constant, y is the effect provoked by [ac, ch.] expressed as
percentages of the maximal effect which can be provoked by very
large amounts of ac. ch. The action of many inhibitors of enzymes
has earlier been shown to follow this law, and this fact has been
the base of certain theories on the mechanism of action of enzyme
poisons. With regard to ac. ch., Clark makes the same conclu-
sions. He postulates that the point of attack of ac. ch. is located
to certain receptors, which react with the ac. ch. molecules accor-
ding to the law of mass action. The effect of a certain amount of
ac. ch. is proportional to the number of receptors, fixing ac. ch,
molecides, in relation to the total amount of receptors in the organ.
Thus, y, which in Clark’s expression represents the effect of [ac.
ch.], also represents the relative number of receptors which have
fixed ac. ch. molecules, and 100 represents the maximal effect of
ac. ch. as well as the total amount of receptors in the organ.

If one postulates that the ac. ch. receptors of various organs are
identical, it would be possible to explain those differences in sen-
sitiveness to ac. ch. of various organs which remain after the in-
fluence of cholinesterase has been eliminated in such a way
that a less sensitive organ contains a larger number of receptors
than a sensitive organ. According to this view, the various amounts
of ac. ch. which are necessary for producing the same effect on
various organs, e. g. 50 p. c. of the maximal effect, should be di-
rectly proportional to the absolute number of receptors.

The antagonism between ac. ch. and atropine has been subject
to quantitative studies by Clark (1926, 1927) and Gaddum (1937).
Both of them found that the relation between concentration and
effect of ac. ch. in atropinized organs also followed the same
simple law of Hitchcock-Lakomuir (with the difference only
that the value of the constant K is lower after addition of atropine).
On the basis of this fact Gaddum (1937) formulated a mathemat-
ical expression for the relation between concentration and effect
of ac. ch. in the presence of arbitrary amounts of atropine. Clark
(1937) found that this expression was well satisfied by values
experimentally found. Both authors made the following inter-
pretation; atropine and ac. ch. compete for the same receptors;
the decrease in sensitiveness to ac. ch. is caused by the fixation
of atropine at the ac. ch. -receptors, thus blocking the receptors
to ac. ch. At the presence of both atropine and ac. ch. a balance
is brought about between ac. ch.-molecules, atropine-molecules, free
receptors and receptors fixing atropine and ac. ch. respectively.

INFLUENCE OP ATROPINE.

157

This means that if a certain effect of ac. ch., c. g. 50 p. c. of the
maximal effect, is to be diminished to another fixed effect, e. g.
25 p. c. of the maximal effect, the same number of receptors in rela-
tion to the total rmniber of receptors must be fixing atropine, what-
ever organ is chosen. If one postulates that the receptors of various
organs are identical, the amount of atropine which is necessary
to diminish the effect of ac. ch. from 50 to 25 p. c. of the maximal
effect must be directly proportional to the absolute number of
receptors. As the amount of ac. ch. necessary for producing
the 50 p. c. effect is also directly porportional to the absolute
number of receptors, the amounts of atropine and ac. ch. in question
must be directly proportional to each other, i. e. the ratio between
the amount of ac. ch. provoking the 50 p. c. effect and the amount
of atropine that lowers this effect to 25 p. c. should be constant,
whatever organ is chosen.

To test the validity of this discussion I have made determina-
tions of these amounts of ac. ch. and atropine on a series of organs
of frog. JIuscarinelike as well as nicotinelike effects were examined.
The ratio between these amounts of ac. ch. and atropine was found
to be fairly constant.

Because of these results I made some experiments on the influ-
ence of atropine on “quick contractions” of skeletal muscles
provoked by intraarterial injections of ac. ch. and on the trans-
mission between motor nerve and skeletal muscle.

A. The quantitative relation between ac. cli. and
atropine in various organs.

Method.

The experiments were made on heart, stomach, intestine, and a
series of skeletal muscles of frog {Bana lemporaria). In all organs
the influence of the esterase was abolished by adding eserine sulphate
1 ; 200 000 to the Ringer solution.

The hearts were suspended according to Straub and fed with oxy-
genated Ringer solution. As it was noted that the hearts became more
sensitive to ac. ch. when they grew weaker, the hearts were used only
as long as the beats had their original heights as registered by an isotonic
lever. To avoid anoxia the Straub cannulas were provided with specially
short and wide necks. As known, the proportion between the inotropic
and the chronotropic effect of ac. ch. and some other inhibiting drugs
varies greatly. Bvenduringtheveryinfluenceofac.ch. a sudden change
in this proportion is often noted. RTien the inotropic or the chrono-

158

N.-O. ABDON.

tropic effect is separately used as a criterion on the magnitude of
ac. ch. action — as in these experiments — these conditions may cause
a rather great error. Therefore, only those experiments were considered
where the hearts reacted solely inotropically or chronotropically.

The other organs were suspended in oxygenated Ringer solution in
the ordinary way. The shortenings of the muscles were registered with
isotonic levers. In every experiment the height of the maximal ac. ch.-
effect was first determined by adding an amount of ac. ch. which
was 1,000 times larger than the treshold concentration. The maximal
shortening was determined again at the end of every experiment and
if the values were not identical the experiment was ignored. The
amount of ac. ch. which provoked 50 p. c. of the maximal effect was
carefully determined. The amount of atropine sulphate was titrated
which was necessary to diminish the effect of ac. ch. from 50 to 25 p. c.
of the maximal effect. As mentioned below, atropine takes such a
considerable time to develop its full antagonistic action on skeletal
muscles that it was not considered proper to test more than two con-
centrations of atropine on each preparation of skeletal muscles. If
the second test was not successful the experiment was not carried on
any further.

At the determination of the magnitude of ac. ch.-action it is usual
to measure the height of muscle shortening 3 to 5 minutes after the
addition of ac. ch. In the present experiments the added amount of
ac. ch. was allowed to act until the curve of shortening became asymp-
totic. In the. case of skeletal muscle this ensued after 5 — 10 minutes
and in the case of stomach or intestine preparations after 3 — 5 minutes.
The velocity of development of the contracture of the individual pre-
paration was as a general rule constant, brit sometimes it was observed
that the velocity changed greatly during the course of experiment.
In spite of this change, the same amount of ac. ch. gave the same
definitive height as before.

In these experiments I bad to determine the effect of a certain
amount of atropine after it had developed its antagonistic action
completely. Riessbr and Neuschloss (1922) noted that atropine
developed its action on skeletal muscles more slowly than ac. ch.,
but the literature does not seem to offer any definite statements
as to the time required for the complete development of the
antagonistic effect. Most observers have arbitrarily treated the
preparations with atropine for 5 or 10 minutes before the addition
of ac. ch. On preparations of stomach or intestine of frog this
time seems to be sufficient, while skeletal muscles have to be
exposed to atropine for a considerably longer period before full
effect of a certain dose of atropine is achieved. Even in case of
a thin muscle as m. rectus abdominis of frog the necessary period
of time was found to be about 50 minutes (see fig. 1). This was
determined by titrating the amount of ac. ch. necessary for

INFLUENCE OP 4TB0PINE.

159

Fig. 1. The development of the ae. ch.-antagonistic action of atropine on m.
rectus of frog ns a. function of time.

X = time in minutes after the addition of atropine

y = the increase of the amount of ae. ch. necessarj’ for provoking 60 p. c. of
the maximal effect. (Calculated as percentages.)

The upper curve represents the action of atrophine sulphate 1 : 60,000, the
middle curve represents the action of atropine sulphate 1 : 120,000, and the lower
curve the action of atropine sulphate 1 : 250,000.

provoking a 00 p. c. effect before and at varying intervals after
the addition of atropine. In consequence of this finding the
skeletal muscles used in the followung experiments were exposed
to atropine during at least 60 minutes.

Resttlts.

As shovn by fig. 2, the amounts of ac. ch. required for giving
50 p. c. of the maximal effect on the different organs vary vrithin
nearly 3 potvers of 10; from 1 : 200 millions in the case of frog’s
heart to 1 : 250,000 in the case of m. dorsalis scapulae. As to
the hearts, the sensitiveness found in these experiments is in
accordance Tvith values found by Cl.4EK (1926) and Beznak
(1934); as to the skeletal muscles the values agree ivith those given
by B^'achholder and Ledebcr (1930). The amounts of atropine
required for diminishing the effect of ac. ch. from 50 p, c. to 25
p. c, of the maximal effect varied on the whole in relation to the

160

N.-O. ABDON.

Fig. 2 shows tho relation between the amount of ao. eh. Cl necessary for pro-
's^ voking 50 p. c. of the maximal effect in various organs and the amounts of atropine

sulphate which dimmish the 50 p. c. effect to 25 p. c. of the maximal effect.
X = the logarithm of the concentration of ac. ch. Cl calculated as y per ml of
suspension fluid.

y = the logarithm of the concentration of atropine sulphate as y per ml of the
suspension fluid.

H = chronotropic effect on heart
Q = inotropic effect on heart
A = longitudinal muscles of small intestine
▼ = longitudinal muscles of stomach
® = m. rectus abdominis
10^ = m. pectoralis abdominis
Q = m. sartorius
0 = m. gastrocnemius

INFLUENCE OP ATKOPINE.

161

necessary amount; of atropine. Although the values of the ration
between the amounts of atropine and ac. ch. vary within rather
wide limits, the correlation between the amounts in question is
beyond doubt, the correlation coefficient being + 0.91 J;; 0.02G.
Thus the expression presented in the introduction is satisfied:

(ac. ch.) ^ ^

(atropine)

Thus, it may be said that those properties of an organ which
make large amounts of ac. ch. necessary for provoking an effect,
obviously also make large amounts of atropine necessary for the
demonstration of the antagonism, whether the antagonism is
studied on muscarinelike or nicotinelike effects. The quantitative
differences between nicotinelike and muscarinelike actions with
regard to the atropine antagonism do not constitute any real
distinction between the actions of ac. ch. The behaviour towards
atropine rather augments those similarities between the two
actions of ac. ch. which are signified by the uniformity in the
influence of many potentiating and antagonistic agents.

B. The antagonistic action of atropine on the “quick
ac. ch.-contraction”.

As mentioned in the introduction of this paper, it has been
shown that the action of ac. ch. on skeletal muscles is abolished
by atropine if the muscles are suspended and ac. ch. added to
the suspension fluid. The effect of intraarterially injected ac. ch
on skeletal muscles of frog or rabbit is, however, not found to be
antagonized. Dale and his co-workers have offered an explana-
tion of this phenomenon that agrees with the theory of humoral
transmission. They suppose that atropine forms a barrier sur-
rounding the receptors, which blocks the way to the ac. ch.-mole-
culcs which diffuse into the muscle from the suspension fluid.
When injected intraarterially, however, the ac. ch. comes into
such an intimate relation to the receptors that the atropine
barrier need not be passed. The ac. ch. that is liberated at the
stimulation of motor nerves of striped muscles also comes into
that intimate relation to the receptors and is therefore not an-
tagonized by atropine, while ac, ch. liberated by the heart vagus
acts after diffusion into the surrounding tissues and is antagonized.

162

N.-O. ABDON.

It should be noted, however, that this view on the mode of action
of atropine does not concord with the quantitative studies of the
antagonism between ac. ch. and atropine, which make it probable
that atropine acts on the very ac. ch.-receptors.

In the experiments on suspended muscles rather large amounts
of atropine have been used. In the case of suspended mammalian
muscles Dale and Gaddum (1930) found the effect of ac. ch. to
be abolished by atropine sulphate 1 : 3,000 while atropine sul-
phate 1 : 150,000 only diminished the effect of ac. ch. In the
in vivo experiments on “quick contractions” atropine was injected
intravenously in such amounts that the actual concentration of
atropine in the muscles must have been much smaller than in
the suspended muscles. In most “quick contraction” experiments,
amounts of atropine were used, which were just sufficient to abol-
ish the effect of ac. ch. on heart and blood pressure; in a few
experiments 10 — 20 times larger amounts were used. From the
data in fig. 2 it is evident that still larger amounts should have been
used. At the calculation of the concentration of atropine in the
muscles from the intravenously injected amount it is also necessary
to remember that only a slight part of the injected atropine is
fixed in the muscles; the main part is fixed by the liver and the
kidneys (Oelkers, Haetz and Binteln, 1932). It therefore
seems possible that the amazing difference between ac. ch. ad-
ministered to the suspension fluid and ac. ch. administered intra-
venously with regard to the susceptibility to atropine is due only
to the fact that greatly' different concentrations of atropine were
used in both cases.

Therefore, I have made some experiments on the influence of
larger amounts of atropine on the effect of intraarterially injected
ac. ch. The experiments were made on m. gastrocnemius of frog
and rabbit. In the experiments on frog the gastrocnemius was
suspended in Kinger solution and the atropine added to the
suspension fluid. In the experiments on rabbits the atropine was
injected intravenously. Independent of the way of administration
of atropine the “qiiick contractions” were abolished. The prepara-
tions for intraarterial injections were made according to Bbown
(1937) and Brown, Dale and Feldberg (1936); as to the details
of the technique these communications are referred to.

Experiment 3. The influence of atropine on the “quick con-
tractions” of frog’s gastrocnemius.

INFL'UENCE OE ATROPINE.

163

On a large Hungarian rana esculenta the sciatic artery was prepared
for injection according to the »close arterial method)). The proximal
end of m. gastrocnemius was fixed by a pin, stuck through the femur
into a small wooden plate, which was then placed in a Petri dish filled
with oxygenated Binger solution. Tendo achilleus was connected to an
isometric lever in such a way that m. gastrocnemius was kept immersed
in the Binger solution while the other parts of the frog were not im-
mersed. In these experiments the muscle was not perfused, the oxygen
being supplied only through the suspension fluid.

After a series of contractions, provoked by stimulating the sciatic
nerve with, condenser discharges, 0.2 y of ac. ch. Cl dissolved
in 0.2 ml of Binger solution was injected into the artery. A “quick
contraction” was provoked (see fig. 3 A). Atropine sulphate was

A B

Big. 3. The influence of atropine on the “quick contraction” of frog’s gastro-
cnemius. At -f- intraarterial injection of 0.2 / of ac. ch. Cl. A = before the addi-
tion of atropme, B = 45 minutes after the muscle had been immersed in atropine

sulphate 1 : 10,000.

added to the suspension fluid to a final concentration of 1 ; 10,000.
After being exposed to the atropme for 45 minutes the muscle
could still be stimulated through the nerve, but an injection of
0.2 y of ac. ch. Cl into the artery had no effect (see fig. 3 B). Even
25 7 of ac. ch. Cl dissolved in 0.2 ml of Binger solution gave no
contraction of m. gastrocnemius, although this amount was large
enough to cause a distant effect, i. e. a contracture of mm. recti
abdomini and of muscles of the foreleg, which had not been
exposed to atropine.

164

N.-O. ABDON.

Experiment 4. The influence of atropine on the “quick contrac-
tions” of the rabbit’s gastrocnemius.

In these experiments the atropine was given by intravenous
injection. As it was a priori probable that so large amounts of
atropine would be necessary that they would interfere with the
brain, the experiments were made on decapitated animals. 4
experiments were made, which gave the same results; one of the
experiments is related below.

A rabbit, weighing 1.8 kilos, ws anaestethized with ether. A cannula
was inserted in the trachea and artificial respiration was given. Decapi-
tation was made by means of the apparatus used in Heymans’ labora-
tory (1932), then no more ether was given. M. gastrocnemius was pre-
pared for intraarterial injection of ac. ch. with maintenance of the
natural circulation according to Brown, Dale, and Deldberg (1936).
The contractions were recorded by means of an isometric lever. For
injection of atropine a caimula was inserted into the femoral vein of
the other leg.

After a series of contractions, provoked by stimulation of the
sciatic nerve with condenser discharges, 10 y of ac. ch. Cl were
injected into the artery, which provoked a “quick contraction”

Fig. 4. The influence of atropine on the “quick contraction” of rabbit’s gastro-
cnemius. At -p intraarterial injection of 10 v of ac. ch. Cl. A = before administra-
tion of atropine. B = 15 minutes after an, intravenous injection of 10 mg of
atropine sulphate per kilo of body weight, 0 = 15 minutes after injection of another
10 mg of atropine sulphate per kilo of bodyweight.

INFLUENCE OF ATROPINE.

165

of great tension (see fig. 4 A). 18 mg of atropine sulphate (10
mg per kilo bodyAveight) were given intravenously and after 15
minutes the injection of ac. cli. was repeated. It still provoked
a ‘contraction”, the tension was, however, considerably smaller
(see fig. 4 JB). Anotlier 18 mg of atropine sulphate were injected
and after 15 minutes 10 y of ac. ch. Cl were injected into the ar-
tery, this rime without any effect (see fig. 4 C). Then further
amounts of atropine were injected, eacli time 18 mg with an in-
torval of 15 minutes between every injection. After injection of
a total amount of 72 mg of atropine sulphate the heart rate was
markedly slower and the arterial blood pressure was diminished.
After the injection of a total amount of 108 mg the heart ceased
to beat, but even now electrical stimulation of the sciatic nerve
provoked contractions.

C. The so called ‘^curare-action” of atropine.

In the experiments on the influence of atropine on the “quick
ac. ch.-contractions” it was alwa}'B observed that while abolishing
the effect of ac. ch., atropine had no influence on the excitability
of the muscle through its nerve. The effect of nerve stimulation
was not even influenced by amounts of atropine 4 or 5 times
larger than those which abolished the ac. ch.-effect.

To judge from these experiments there seems to exist a certain
difference between the effect of ac. ch. and the effect of motor
nerve stimulation with regard to their susceptibility to atropine.
Premous experiments have, however, showir that large amounts
of atropine block the transmission between motor nerve and
skeletal muscle (Bodkin, 18G2, et al.), and some authors count
atropine among substances vdth curare-action [vide Cushny,
1924). As shown by the experiment below — which is in accord-
ance with data given by Cushny (1903) and Haffner (1918) —
it is necessary to use rather massive concentrations of atropine to
demonstrate the “curare-action”, and as emphazised below, this
“curare-action” is not of the same nature as the ac. ch.-antago-
nizing effect to atropine.

Experiment 5.

8 nerve-muscle preparations of m. gastrocnemius of frog were made
in the usual manner. Two preparations were placed in oxygenated
Kinger solution, two in atropine sulphate 1 ; 250, two in atropine

166

N .- O . ABDON.

sulphate 1 : 125, and two in atropine sulphate 1 : 75. Every 30 minutes
the indirect and direct excitability of the muscles were tested by means
of condenser discharges. The result is seen from the scheme below
(+ = contraction, — = no contraction).

honrs

2.5

3.5

4.5

5.5

6.5

directly

-f

-i-

-1-

-I-

-f

-t-

-f

indirectly

-1-

-i-

-f

-i-

atropine

directly

_l_

-1-

-f

-1-

-U

1:250

indirectly

atropine

directly

-i-

-t-

■¥

-f

—

1 : 125

indirectly

-f

—

atropine

directly

•f

—

1:75

indirectly

—

Like similar experiments by other authors, this experiment
shows that skeletal muscles, suspended in rather concentrated
solutions of atropine, lose their indirect excitability after an inter-
val, the length of which is depending on the concentration of atro-
pine. If atropine is- allowed to act on the muscles some further
time, the muscles do nor answer even to direct stimulation.
Haffner (1918) also studied the curare-action of atropine,
added to perfusion fluid. He found that atropine administered
this way exerted the “curare-effect” in smaller concentrations,
1 ; 1,000 — 1 : 2,000. Also the direct excitability was abolished
after prolonged perfusion.

At those concentrations, which are necessary for provoking the
“curare-action” of atropine, many drugs will most probably
block the transmission between nerve and skeletal muscle. A
■priori it seems to be likely that this effect of atropine is unspecific.
With regard to the theory of humoral transmission it is, however,
interesting co stare that the “curare-action” of atropine is not
of the same nature as the ac. ch. -antagonizing effect; it cannot
qe explained as an antagonizing of ac. ch. liberated at the stimula-
tion of the motor nerve.

As early as 1903 Cushny emphazised an essential difference
between the vagomimetic and the “curare-like” action of atropine.
On various organs he studied the action of the two optical isomeres
of hyscyamine, which are the constituents of atropine. He found
that the parasympathetic effect of atropine was most probably
entirely due to its content of 1-hyoscyamine. On the parasympa-
thetic innervation of iris, salivary glands, and heart 1-hyoscya-
mine was found to be twice as effective as atropine and 12 — 15
times as effective as d-hyoscyamine. Laidlaw (1909), probably

INFLUENCE OF ATROPINE. 167

using a purer preparation of d-liyoscyainine, found a stiJ] greater
difference. As to the curare-like effect, however, Cushny found
d-hyscyamiue to be quite as effective as 1-hyoscyamine. For
completeness, I have measured the antagonistic power of both
isomeres (delivered by Burroughs and Wellcome) on the action
of ac. ch. on heart ns well as on m. rectus abdominis of frog. In
both cases 1-hyoscyamine was found to be 30—40 times as effec-
tive as the dextrogjte isomere. Together with the findings of
Cushny, this shows that the “curare-action” is not of the same
nature as the ac. ch. -antagonizing, vagomimetic effect of atropine.

This conclusion is also supported by the fact that the direct
excitability of the muscles is suppressed by atropine, and, according
to Dale and his co-workers the direct excitability has nothing
to do with the liberation of ac. ch. The long period of latency also
speaks in the same direction. In case of atropine sulphate 1:75
the time of latency was found to be more than two hours. As
shown by fig. 1, also the ac. ch.-antagonistic effect has a consider-
able period of latency before a certain amount of atropine has
developed its full action, but on the other hand the greatest part
of the antagonistic effect appears within about 15 minutes.

Thus it must be said that the effect of ac. ch. added to a striated
muscle can be abolished by atropine, while even a massive con-
centration of atropine does not exert any effect on the indirect
excitability of skeletal muscles which can reasonably be explained
as an antagonizing of ac. ch. This discrepancy scarcely accords
ts*ith the theory of ac. ch. as the humoral transmitter at the
myoneural junctions of skeletal muscles.

Dale and his co-workers are of opinion that this discrepancy
cannot be considered to have any decisive significance to the
question of ac. ch. as myoneural transmitter. They point out
that all parasympathetic nerves most probably have the same
transmitter and this transmitter is most probably ac. ch. They
refer to statements in the literature that in certain organs the effets
of stimulation of parasympathetic nerves cannot be antagonized
by atropine, which, however, antagonizes the effect of added ac. ch.
According to their opinion, ac. ch. may be the transmitter at
the myoneural junctions of skeletal muscles, although atropine
exerts no specific action on the indirect excitability of striated
muscles.

Against this argumentation may be said that in reality atropine
suppresses all effects of stimulating parasympathetic nerves, even

168

N.-O. ABDON.

if in some cases it is necessary to use rather large amounts. As
CuSHNY (1924) points out, most observers have only studied the
influence of smaller amounts of atropine, and therefore it is often
stated that atropine has no influence on the effects of stimulation
of the following parasympathetic nerves: vasodilatating nerves of
chorda lingualis, motor innervation of stomach, intestine, urinary
bladder, and of uterus. They are, however, paralyzed by larger
amoimts of atropine. As early as 1895 Langley and Aj^derson
found that about 10 mg per Irilo of atropine paralyzed the motor
innervation of rabbit’s intestine, and in 1896 the same authors
showed that atropine paralyzed the motor innervation of the
urinary bladder. They have been confirmed by later authors.
Eecently Harrison and McSwiney (1936) found the motor inner-
vation of stomach to be paralyzed. Just as is the case mth added
ac. ch., the amounts of atropine necessary for abolishing the effects
of parasympathetic nerves in various organs vary within rather
wide limits. V. E. Henderson (1923) measured the quantities of
atropine necessary for paralyzing various parasympathetic nerves,
which he found to be influenced by increasing concentrations of
atropine in the following order: heart vagus, glandular nerves of
chorda lingualis, vasodilatating nerves of chorda lingualis, and
motor nerve of intestine; the parasympathetic nerves of uterus
were not quite suppressed by the amounts used by Henderson.

It must, therefore, be said that the theory of humoral trans-
mission at present offers no acceptable explanation of the discrep-
ancy between the influence of atropine on the ac. ch.-effect
and on the indirect excitability of skeletal muscles.

Summary.

1) Dale’s original definition of the different actions of ac. ch.
stated that muscarinelike actions are those which are abolished
by atropine, while the nicotinelike actions are not antagonized.
Later experiments show that both kinds of actions are antagonized
by atropine, even if several exceptions from this rule are described
There is, however, a quantitative difference between muscarine-
like and nicotinelike actions with regard to the atropine antago-
nism. As a rule, muscarinelike actions are inhibited by smaller
concentrations, while nicotinelike actions require higher concen-
trations of atropine. This quantitative difference is by some
authors looked upon as a criterion of a fundamental distinction

IJCFLCENCE OF ATHOPIKE.

169

between tbe two kinds of action. The nicotinelike actions require,
liowever, not only larger amounts of atropine, but also large
amounts of ac. ch. It seems possible that those properties of an
organ which make large amounts of ac. ch. necessary also make it
necessary to use large amounts of atropine before the antago-
nism can be demonstrated. It is shown in the present paper that
there is a constant ratio between the amounts of atropine which
arc necessary to provoke a certain effect on various organs of frog
and the amount of atropine which is necessary for diminishing
this effect to a certain degree. Thus, the quantitative differences
between the two kinds of action with regard to the atropine
antagonism do not seem to constitute any real distinction.

*2) Striped muscles of frog must be exposed to atropine during
50 to 60 minutes, before the atropine exerts its full antagonistic
power.

3) Is has been shown previously that atropine antagonizes the
effect of nc. ch. on striped muscles, when these are suspended in
a bath, while the effect of intraarterially injected ac. ch. is not
abolished. In the latter case too small amounts of atropine seem
to have been used. In the present paper is shown that atropine
prevents the “quick contractions” of skeletal muscles of frog or
rabbit, provoked by intraarterial injections of ac. ch.

1) Those amounts of atropine which abolish the “quick con-
tractions” do not have any influence on the excitability of the
muscles through their nerves. If, however, the muscles are
suspended in very strong concentrations of atropine they lose at
first their indirect excitability and then also their direct excita-
bility. Several reasons — {nlcr alia the effect of d- and 1-hyoscya-
mine — show that this so-called ’curare-action” of atropine is
not of the same nature as the ac. ch. -antagonizing effects of
atropine. Thus, while the effects of added ac. ch. is abolished,
atropine exerts no action on the indirect excitability which can
be explained as an antagonizing of ac. ch. It is emphasized that
the theory of humoral transmission at present offers no acceptable
explanation of this discrepancy.

References.

AuumCxV G., Skand. Arch. Physiol. 1929, 58, 1.
Axdrus, E. C., J. Physiol. 1924, 59, 361.

Bacq Z. M. and G. L. Brown, ibidem 1937, 89, 45.

12 — 'i01323. Acta vhys. Scandinav. Vol.I.

170

N.-O. ABDON.

Beznak, a. B. I., ibidem 1934, 82, 129.

Bodkin, S., Arcb. path. Anat. Physiol. 1862, 83, 24.

Brown, G. L., J. Physiol. 1937, 89, 220.

Brown, G. L., H. H. Dale, and AV. Feldberg, ibidem. 1936, 87, 394.
Cannon, W. B. and A. Rosenbldth, Amer. J. Physiol. 1932, 116, 408.
Clark, A. J., J. Physiol. 1926, 61, 530.

— , ibidem 1927, 64, 123.

— Heffters Handbuch exp. Pharm. 1937, Erg. bd IV, 1 — ^223.
Cushny, a. R., j. Physiol. 1903, 30, 176.

— , HeHters Handbuch ex. Pharm. 1924, II: 2 A, 599 — 655.

Dale, H. H., J. Pharmacol. 1914, 6, 147.

Dale, H.H . and J. H. Gaddum, J. Physiol. 1930, 70, 109.

Frank, E., M. Nothjiann and H. Hirsch-Kauefjiann, Klin. Wschr.
1922, I: 37, 1820.

Gaddum, j. H., j. Physiol. 1937, 89, 7 P.

Haefner, F., Arch. int. pharmacodyn. 1918, 24, 547.

Harrison, J. S. and B. A. M. McSwiney, J. Physiol. 1936, 87. 79.
Henderson, V. E., J. Pharmacol. 1923, 21, 99.

Heyjians, C., j. j. Bouckaert and P. Regniers, Le sinus carotidien,
Paris, 1933.

JuLLiEN, A. and G. Morin, C. r. Soc. Biol. Paris, 1931, 106, 187.
Kahlson, G. and B. Uvnas, Skand. Arch. Physiol. 1938, 78, 40.
Laidlaw, P. P., j. chem. soc. 1909, 99, 1966, quoted from Cushny, 1924.
Langley, J. N. and H. K. Andersson, J. Physiol. 1895, 18, 67.

— , — , ibidem, 1896, 20, 372.

Minz, B., Arch. exp. Path. Pharmak. 1932, 168, 292.

Oelkers, H. a., VI. Haetz and K. Rinteln, Arch. Pharmazie, 1932,
270, 520, quoted from Heffter’s Handbuch exp. Pharm. 1936,
Eg. bd. Ill, pp. 24 — ^25.

Raventos, j., j. Physiol. 1936/37, 88, 5P.

Riesser, O. and S. M. Heuschloss, Arch. exp. Path. Pharmak. 1921,
91, 342.

Rosenblueth, a., Amer. J. Physiol. 1932, 100, 443.

Wachholder, K. and J. v. Ledebur, Pfliig. Arch. ges. Physiol. 1930,
225, 627.

AVahlquist, B., Skand. Arch. Physiol. 1935, 72, 133.

From the Institute of theoretical Physics and the Zoophysiological
Laboratory, University of Copenhagen.

Rate of Penetration of Pliospliate
into Muscle Cells."

G. HEVESY and O. REBBE.

(With 1 fig. in the text.)

Certain constituents of the voluntary muscle are able to diffuse
out of the muscle into a surrounding saline and can also diffuse
into the muscle if previously dissolved in the saline in a suffi-
ciently high concentration. There exists, in such cases, a given
concentration of the substance in saline which will be in equilib-
rium with the tissue. This critical concentration provides a
measure for the concentration of the substance in the tissue
or rather in such part of the tissue as is concerned in the diffusion.
M. G. Eggleton (1933) carried out such experiments- in respect
of phosphate exchange by immersing in Ringer’s fluid an excised
frog muscle at 2° for a few hours and found that in the resting
muscle 20 — 30 per cent of the muscle water, corresponding to
about 16 — 24 per cent of the weight of the fresh muscle, was in-
volved in the diffusion system. Since 8 — 16 j)er cent of the weight
of the gastrocnemius is composed of the interspaces into which
the ■ phosphate of Ringer’s fluid will easily penetrate, the result
mentioned above suggests either that phosphate can diffuse only
into a certain fraction of the tissue beyond the extracellular
volume or else tliat phosphate ions can penetrate onl} slovly
into the muscle cells. That the latter alternative is more likely
follows from the fact that no perceptible decrease in the total
acid soluble phosphate content of muscles is apparent after fatigue,
though a very perceptible increase in the inorganic P content

' Received 5 September 1940.

' Comp, also Stella (1928).

172

G. HBVESr AND 0. EEBBE

of such, muscles takes place. If the cell membranes were easily
permeable to phosphate ions, a part of this excess phosphate
should soon leak out into the plasma.

The application of the method of isotopic indicators permits
us to follow the path of the labelled phosphate ions introduced
into the circulation by making use of radioactive measurements.
Due to the great sensitivity of this method it is possible to deter-
mine even very small amounts of labelled phosphate ions which
migrate imder strictly physiological conditions into the muscle
cells during a few hours or less.

Description of the method.

Sodium phosphate of negligible weight containing radioactive
P as an indicator is introduced into the circulation of the frog
(injected into the lymph sack). After the lapse of, for example,
ten hours, we compare the labelled phosphorus (®^P) content of a
plasma sample and of a gastrocnemius sample of the same weight
by determining their radioactivity. Let us assume that 1 gm
plasma is found to be 5 times more active than 1 gm muscle tissue
and the interspaces to make up of the muscle’s weight, then
the amount of phosphate ions which penetrated from 1 gm plasma
into the cells of 1 gm muscle works out to be of that present

in 1 gm plasma or, in general, is — — ip, where m denotes the

activity of 1 gm muscle, p the acti\dty of 1 gm plasma, and i the
size of the interspaces as a fraction of the muscle weight. When
carrying out the calculation given above, we assume that ®*P
becomes equally distributed between plasma and interspaces in
an early stage of the experiment. How far this assumption is
justified will be discussed later.

The experiment described above is carried out under strictly
physiological conditions; the phosphorus content of the plasma
and the muscle remains practically constant during the experi-
ment and we can, therefore, conclude that the penetration of
phosphate ions from the plasma into the muscle cells was followed
by a migration of an equal number of phosphate ions from the
muscle cells into the plasma. If an equilibrium is reached with an
uptake of less than corresponding to the total water content
of the system, we must assume saturation of a certain fraction
of the tissue but, when the relative concentration of ’-P conti-

RATK OF PENETRATION OF PHOSPHATE INTO MUSCLE CELLS. 173

nuously increases in the muscle, we can utilize the results to mea-
sure the rate of exchange between cellular and extracellular
phosphorus.

The application of the method outlined above requires the
hnowledge of the extent of the interspaces of the muscle tissue.
The size of the interspaces can be obtained by comparing the
chloride or sodium content of muscle and plasma samples of the
same weight or by other methods. When applying the first
mentioned method, the assumption is made that all sodium and
chlorine present in the tissue is to be found in the interspaces.
U'enn and Cobb (1935) state for the chlorine space of the sartorius
of rana pipiens values var}’ing between 7.5 and 16.0 per cent,
the average being 11.3 per cent.

To determine the size of the extracellular space of the gas-
trocnemius of the Hungarian frog (Eana esculenta) used in our
experiments, we administered labelled sodium along tvith the
labelled phosphate. By measuring the distribution of the labelled
sodium between plasma and fresh gastrocnemius of equal
weight we arrive at a figure indicating the extracellular volume
of the muscle. The amount of extracellular phase (E) in percent
of the weight of the muscle is calculated (Manerv and Hastings
1939) from the equation

^ (^^Ha)„,» 0.97 -100

(«Na)p'0.99

in which the subscripts m and p represent muscle and plasma,
respectively^.

The size of the interspaces being known, the measurement of
the distribution of between plasma and muscle permits us,
as described above, to determine the amount of which pene-
trates into the muscle cells.

A simultaneous measurement of the radioactivity of sodium
and phosphorus is made possible by the fact that ^^Na decays
with a half-life period of 14:. 8 hours while ®”P decays with a period
of 14.5 days. S\Tien measuring the activity of the sample, for
example, two weeks after the start of the experiment, the -’Na

7 IVlien' carrying out the calculation mentioned above ve assume that the extra-
cellular phase is identical with the ultrafiltrate of serum. The vater conten^f
the extracellular phase is assumed to he 99, that of the plasma 95 per cent. We
arrive at the figure 0.97 by taking into consideration that the sodium ion con?
centration of the plasma and its ultrafiltrate somewhat differs and by calculating
the difference from the Gibbs-Donnan equation.

174

G. HEVESY AND 0. RBBBE.

originally present in the tissue and the plasma is entirely decayed
and the activity measured is solely due to the content. Let us
say Ave measured at that date 10 counts per minute, then two weeks
previously the activity of the of the preparation was 20. counts.
Assuming Ave measured, two weeks preA'iously, a total of 100
counts, then out of these 80 counts Avere due to the ^^Na content
and 20 counts to the ®^P content of the gastrocnemius sample.
The accuracy of the determination can be augmented by ad-
ministering a preparation which is showing a strong ^*Na and a
comparatively weak ®^P actmty. In view of the variability of the
size of the interspaces in different muscles and in different frogs
(comp. Eggleton et alia 1937), it can be of importance to deter-
mine the extracellular volume of the muscle the permeability of
which to phosphate ions is to be determined.

In experiments of short duration, the P actmty of the plasma
is solely due to the presence of radioactive inorganic P, the
amount of radioactive phosphatides present being negligible and
the plasma containing but an insignificant amount of acid soluble
organic P. In experiments of very long duration, we are not
permitted to use the total acti\’ity of the plasma, but we have
to extract the plasma inorganic P and to compare its activity
Avith the total activity of the muscle cells, assuming that only,
inorganic P can penetrate into the cells and is afterwards largely
combined in the cells.

Results.

The distribution of ^*Na and of ®*P between equal weights of
plasma and gastrocnemius muscle is seen in Table 1. The ^*Na
was administered as 0.6 per cent sodium chloride solution which
contained a negligible amount of active sodium phosphate. The
solution was injected into the lymph sack of the frog. While, in
experiments taking no more than two hours, the phosphate solution
was administered at the start of the experiment, in experiments of
longer duration a steadily decreasing volume of the labelled P
solution was injected at intervals of two hours all through the
experiment. Due to the uptake of the labelled phosphorus by
bone and other tissue, the ®^P concentration of the plasma strongly
decreases during an experiment unless kept up in this way. In
the case of sodium which is not taken up by the cells to any
appreciable extent, the ‘“‘Na concentration of the plasma does not
much decrease with time.

HATE OF PENETRATION OP PHOSPHATE INTO MUSCLE CELLS. 175

In Table 1 it is seen that, while the apparent sodium apace makes
out 4 per cent only of the weight of the muscle after the lapse of
3 min., after 20 minutes 14 per cent are found, which almost
corresponds to the actual sodium space. The fact that between
20 minutes and 4 days the ratio of the ^^Na content of plasma
and muscle of equal weight hardly changes is showing clearly
that no significant uptake of 2 =*Na by the muscle cells takes place
in the course of 4 days.

Table 1.

Distribution of ~'^Na and ^~P, respectively, between plasma and
gastrocnemius^ of equal weight at 22°.

Time after
administration of the
“Na and ®-P

--Nap

92p

Sip

3 min

0.04G .

0.016

1 9 min

0.12

—

20 min

0.14

—

1 hour

0.166

10 hours

0.16

0.29

2 days

—

1.19

4 days

0.14

2.47

The Bubscripts m and p represent muscle and plasma, respectively.

The amount of which penetrates into the muscle tissue in
the course of the first 3 minutes is much smaller than that of
^*Na; this phenomenon is due to a slower diffusion of the phosphate
ions through the capillary wall. After the lapse of 1 hour, a slightly
greater percentage of the plasma®*P is found in the muscle tissue
than of the plasma **Na, the difference increasing with time.
This phenomenon is due to a successive penetration of ®“P into
the muscle cells. The amount migrated into the cells is obtained
by subtracting the sodium space from the apparent phosphorus
space. After 10 hours we find, as seen in Table 1 and Pig. 1, that
the amount of ^-P which diffused into the cells of 1 gm gastro-
cnemius makes up 14 per cent of the ®“P content of 1 gm plasma or
about 100 per cent of the ®®P present in the interspaces of 1 gm
gastrocnemius. After the lapse of 4 days, the corresponding figures
are 233 and 1653 per cent, respectively. The rate of penetration

* The ertracelMar volume in per cent of^the muscle vreight is obtained by
multiplying the above figures by 98 (see p. 173).

176

Q. HEVESr AND 0. BEBBB.

of phosphate ions into the muscle cells of the frog is, thus, a very
slow one even at 22° and still slower at lower temperature. After
the lapse of 10 hours at 0°, the apparent phosphorus space of the
gastrocnemius was found to be 18 per cent; thus, only about'Vs
as much labelled phosphate diffused into the cells at 0° than at 22°.

As mentioned above, we arrive at the values stated for the
amount of labelled phosphate v.'hich migrated into the muscle
cells by subtracting from the total amount of found in the
muscle the amount of present in the interspaces. The accuracy
of the figures obtained depends largely upon the accuracy of the
figure assumed for the size of the interspaces.

In determining the extracellular volume we make two assump-

RATE OF PEXETBATIOK OF PHOSPHATE INTO MUSCLE CELLS. 177

lions: a) We assume that all sodium or chlorine present in the
muscle cells is exclusively found in the interspaces; b) we assume
the concentration of sodium and chlorine, respectively, to be the
same in the plasma water and the extracellular fluid. Much
evidence is available that these assumptions are essentiall)’- correct.
It is possible, however, that: a small amount of sodium or chlorine
penetrates into the cells (Hastings and Eichelderger 1937) and
also that the extracellular fluid does not show exactly the same
sodium or chlorine content as the plasma (comp. Manery et alia
1938). The ske of the interspaces calculated from the distribution
figures of sodium and chlorine, respective!}', is, however, aboxit
the same. Fenn and Cobb (1936) found the average sodium space
and the average chlorine space of the rats’ muscle to be 12.6 and
11.4 per cent, respectively. Manery and Hastings (1939) state
rhe apparent e.xtracellular space of the gastrocnemius of the
rabbit to be 11.3 per cent, calculated from the distribution of
chlorine, and 11.0 per cent from the distribution of sodium, while
for the abdominal muscle they give the figures 16.3 and 13.9 per
cent, respectively.

We assume, furthermore, that an equal distribution of the
labelled phosphate between plasma and extracellular fluid takes
place at an early stage in the experiment. This assumption in-
volves some uncertainty. After the lapse of 3 minutes (see Table 1),
an equal distribution is far from being reached by either sodium
or phosphate; after 1 hour equality may be reached also by the
phosphate, hut the possibility cannot be excluded that the equal-
ity of the sodium space and the phosphate space found after the
lapse of 1 hour is a fortuitous one and is due to the fact that
some penetrates into the cells before the equipartition men-
tioned above was obtained, the sum of the cellular and extracellular
P present in 1 gm muscle making out 14 per cent of the ®-P content
of 1 gra plasma. In the case of the brain tissue, the capillaries of
v;hich are only at a slow rate permeable to phosphate, ^ve found
obvious indications of a penetration of ®"P into the cells before an
equipartition of ®*P between plasma and extracellular fluid
was obtained. It is, therefore, of importance to find a method
which permits us to determine the amount of ^’^P penetrating into
the cells without having to make any assumption regarding the
size of the interspaces and the time involved in obtaining an equal
distribution of ®“P between plasma and extracellvdar fluid. Such
a method will be described in the following section.

178

G. HE VEST AND 0. llEBBE.

Description of the modified method.

'When applying the modified method, we compare the activity
of the inorganic P of a plasma sample of known weight with the
activity of the organic phosphorus extracted from the muscle
sample of the same weight. This method is based on the assump-
tion that the organic phosphorus compounds present in the muscle
tissue are formed in the muscle cells from inorganic phosphate
and that, correspondingly, all active P atoms present in the organic
constituents of the muscle are such which passed from the plasma
into the cells as inorganic ®^P. Since some of the active phosphate
penetrated into the cells will not have had opportunity to he
incorporated into organic molecules, but will remain in inorganic
state, the method here outlined will give a lower limit for the ex-
tent of phosphorus exchange between plasma and muscle cells.
By adding to the activity of the organic P of the muscles that of
the cellular inorganic P we arrive at the total cellular activity.

In experiments of several hours’ duration or more, we can
estimate the amount of cellular inorganic ®*P by the following
consideration: Let us consider an experiment taking 10 hours.
We find that, in this experiment, 66 per cent of the ®*P content
of the acid soluble Organic P of the muscle is present as creatine-
phosphoric acid P and that the amount of creatinephosphoric
P makes out 2.4 times that of the inorganic P present as such in
the muscle. The last mentioned data are obtained by the usual
chemical determination of creatine P and inorganic P, respectively.
Since from activity data we know that, in the course of 10 hours,
almost all the creatinephosphate molecules present in the muscle
get renewed, the ®*P content of 1 mgm inorganic P wiU be about
equal to the ®*P content of 1 mgm creatine P, From these daba it
follows that the activity of the cellular inorganic P. makes out
66 : 2.4 = 28 per cent of that of the cellular acid soluble organic
P. We have, thus, to add to the values obtained for the acid
soluble organic cellular ®*P content of the muscle (see Table 3,
column 2) 0.28 times the value obtained in order to ger the value
of the total ®*P migrated into the muscle ceils during 10 hours. In an
analogous way the other figures seen in column 2 of Table 3 were
obtained. The correction due to the presence of ®*P in the cellular
inorganic P fraction w'as smaller in experiments of longer duration.

In the consideration stated above we have disregarded the fact
that the amount of inorganic P extracted from the muscle is partly

RATE OF PENETRATION OP PHOSPHATE INTO MUSCLE CELLS. 179

extracellular P. TMs procedure is permissible when dealing, as
above, vitli chemical magnitudes alone in view of the fact that
the amount of cellular inorganic P is about 60 times larger than
that of the extracellular inorganic P of the muscle. We meet,
however, very different conditions when considering the radio-
actmty of the cellular and extracellular inorganic P, respectively,
(which is not the case in the consideration made above). Due to
the slow migration of the phosphate ion into the muscle cells, the
actmty of 1 mgm extracellular inorganic P may be many hundred
times larger than that of 1 mgm cellular inorganic P.

Tabic 2.

Disiribution of behueen the acid soluble organic constituents
extracted from 1 gm gastrocnemius and the inorganic
phosphate extracted from 1 gm plasma.

Time after
administration
of ==P

Distribution

coefficient

Percentage’ of organic
acid soluble P re-
placed by plasma P

10 hours

0.11

0.6

2 days

1.02

2.4

4 days

2.11

8.0

Table 2 contains data on the ratio of the activity of the in-
organic P of plasma samples and that of the acid soluble organic
P fractions isolated from muscle samples having the same weight
as the plasma samples. These ratios are stated in column 2 and
indicate, as mentioned above, the lower limit of the fraction of
the plasma P which exchanged with cellular acid soluble P during
the experiment. Column 3 contains data on the percentage of the
organic acid soluble P which was replaced by plasma P during
the experiment. In the course of 4 days, 8 per cent was replaced.

The amount of ®®P incorporated into the non acid soluble P of
the frogs’ muscle is very restricted; After the lapse of 4 days, the
number of ®®P atoms incorporated into phosphatides makes up

2.5 per cent of the amount present in the acid soluble compounds.
For the ^^P incorporated into residual (protein) P, the correspond-
ing figure w as found to be 2.6 per cent. Thus, the amount of

^ The ratio of the acid soluble P cqntent of 1 gm muscle and 1 gm plasma
was found in the frog killed after 10 hours, 2 days and 4 days to be 31, 42 and
26.4 respectively. The variation of the above ratio is to a large extent dne to a
variation in the acid soluble P content of the plasma which was found to be 3.6

3.6 and 4.8 mgm per cent respectively. ' ’

180

G. HEVESY ANB 0. BEBBE.

incorporated into all phosphorus fractions present in the muscle
is, after the lapse of 4 days, but 5 per cent larger than the amount
incorporated into the acid soluble compounds. In experiments of
shorter duration the difference is still less.

Table 3.

Distribution of between the total cellular acid soluble phosphate
extracted from 1 gm gastrocnemius and the inorganic
phosphate extracted from 1 gm plasma.

Time after
administration

Distribution coefficient
obtained by tbe modified
method (organic fraction
extracted, share of in-
organic P computed)

Distribution coefficient
obtained by the original
method (direct comparison
of the activity of muscle
and plasma of equal
weights after subtraction
of content of the

extracellular volume) j

10 bours

0.14

0.14 1

2 days

1.20

1.05 i

4 days

2.50

2.33 ;

In Table 3, the distribution coefficient of betYreen cellular P
and plasma P arrived at by the two different methods is given.
Column 2 contains the distribution coefficient calculated from the
®^P content of the organic fractions and of the inorganic phosphate
content of the muscle, as described on p. 178, while in column 3
the results are given which were obtained by comparing the
activity of the muscle tissue with that of the plasma. The results
obtained agree fairly well.

It is of interest to remark that the labelled phosphate was found
to penetrate at a faster rate into the ceUs of mammalian muscle
than in those of the frog (Hahk et alia 1939). The ratio of
the =^*P content of 1 gm gastrocnemius and of 1 gm plasma of the
rabbit was found after the lapse of 4 hours to be 0.6.

Discnssion.

The concentration of the inorganic phosphate present in the
water of the muscle cells is about 100 times larger than that
present in the plasma water. This puzzling difference can be
explained in two different ways. ^Ye can assume that most of
the inorganic phosphortis extracted after a most careful treat-

KATE OF PENETRATION OF PHOSPHATE INTO MUSCLE CELLS. 181

ment of tlie muscle was not present as inorganic phosphate pre-
vious to extraction in the muscle but in the form of a very labile
phosphorus compound. Creatinephosphoric acid is a fairly labile
compormd which can only be extracted without decomposition
if the operation is carried out at a low temperature and a very
fast rate. It is quite conceivable that a decomposition of other
still more labile phosphorus compounds during extraction cannot
be avoided. But the existence of the great difference in the
inorganic phosphate concentration of plasma and cell water can
be explained in a very different wmy as well.

The muscle cells take up when formed a comparatively large
amount of inorganic P. This high inorganic P content is maintained
all through life, the cell walls being impermeable to phosphate
ions or, as far as a restricted permeability is present, the phosphate
lost by the cells is compensated by a secretion of an equal amount
of phosphate from the plasma into the muscle cells. Numerous
examples of such active secretion are reported by Krogh (1939),
and the application of his \'iews to the present problem leads to the
last mentioned explanation.

It is not possible at present to decide which of the explanations
mentioned above is the correct one; the application of radioactive
P as an indicator leads, however, to the result that a restricted
permeability of the muscle cell wall to phosphate ions is actually
present. Professor Krogh has kindly drawn our attention to a
possibility of deciding which of these explanations is the right one.
According to his view, the primary process is the loss of some
cellular phosphate by leakage through the cell wall. The extent
of the active secretion into the cells is that necessary to com-
pensate for the loss by leakage and is determined by the extent
of the latter. Let us increase the phosphate concentration of the
plasma by administering large amounts of phosphate for example.
This increase should, according to the \dew' cited above, not in-
fluence the amount of labelled phosphate penetrating into the
cells wliile, in the case that the entrance of labelled phosphate
into the cells is due to diffusion, the amount entering the cells from
a plasma containing more phosphate should be larger than from
one containing less.

We wish to express our hearty thanks to Professor Niels Bohr
and Professor August Krogh for numerous facilities most kindly
put at our disposal.

182

G. naVESY AND 0. EEBBE.

Summary.

Labelled sodium and labelled pbospbate are injected simul-
taneously into the lymph sack of the frog and the distribution of
the radioactive sodium and the radioactive phosphate between
plasma and muscle of equal weights is determined. A constant
partition ratio of the. radioactive sodium is obtained after the
lapse of about 20 minutes. From this ratio the volume of the
interspaces of the muscle can be calculated. .Tn the case of phos-
phorus, the partition ratio increases even after the lapse of many
days because of continued penetration of the labelled phosphate
into the cells. The difference of the partition ratio of the radio-
active phosphorus and the radioactive sodium permits us to cal-
culate the amounr of which penetrated into the muscle cells.
Since all phosphorus atoms present in the plasma can be assumed
to show the same behaviour as the ^^P atoms we can compute
from the figures obtained the amount of plasma inorganic P which
exchanged with cellular P during the experiment. In the course
of 4 days at 22°, 0.082 mgm P w^as found to penetrate into the
cells of 1 gm gastrocnemius muscle and vice versa. At 0°, of
the above value was found.

An alternative method which is independent of the knowledge
of the size of the extracellular space is based on the determina-
tion of the comparison of the active inorganic phosphorus content
of the plasma with the active organic phosphorus content of the
muscle. When applying this method, one assumes that the organic
phosphorus compounds of the muscle are renewed inside rhe cells.
By this method, the amount of plasma P penetrated into 1 gm
muscle in the course of 4 days was found to be 0.088 mgm.

Eeferences.

Eggleton, M. G., J. Physiol. 1933. 79. 31.

— , and P. Eggleton, J. Physiol. 1937. 90. 167.

Eichelberger, L. and A. B. Hastings, J. Biol. Chem. 1937. 118. 197.
Eenn, W. 0. and D. M. Cobb, Amer. J. Physiol. 1935. 112. 41.

Hahn, L. A., G. Ch. Hevesy and O. H. Eebbe, Biochem. J. 1939.
33. 1549.

Krogh, a., Osmotic Regulation in Aquatic Animals, Cambridge 1939.
Manery, j. F., j. S. Danielsen and A. B. Hastings, J. Biol. Chem.
1938. 124:. 359.

Manery, J. F. and A. B. Hastings, J. Biol. Chem. 1939. 127. 688.
Stella, G., J. Physiol. 1928. 66. 19.

. Aus der Anatomiscben und der Pharmakologischen Abteilung
des Jfaroliniachen Instituts in Stockholm.

iiOer den Eiiifluss lokaler pliysikalisclier
und cliemischer Hantreizc auf die
p eriplierc Bliitv evteihmg.^

Ton

GOSTA von REIS wnd PRITIOF SJOSTRAND.
out 4 Fignren im Text.)

In friiheren Arbeiten (G. VOK Reis und P. Sjostrajjd 1937, 1938)
wurden die Rcsulfcnte eiuer Untcrsuchung iiber den Einfluss von
lokaler Hautreizung mittels Senfol und Ultra tdolettetrablung auf
die pcripherc Blutvcrteilung in der Leber und Nierenrinde bekannt-
gegeben. ^lit bciden Reizmittcln xvurde cine erheblicbe Zunabme
der peripheren Blutmonge in dicsen Organen erzielt, ein Effekt,
tvelcher ausblieb, Venn dcr gereiztc Hautbezirk vorber denerviert
\Yorden var. Die vorliegende Arbeit entbalt einerseits die Ergeb-
nis.se einer Untersuebung iiber die AVirkung lokaler tbermischer
Hautreizung auf die peripbere Blutmonge in der Leber und Nie-
renrinde sowie auf die AnzabI offener Kapillaren im M. masseter,
andererseits die Resultate von Bestimmungen der AnzabI offener
Kapillaren im JL mas.seter bei gelegentlicb der friiber besclrrie-
benen Versuebe mit lokaler Hautreizung mittels Senfol und Ultra-
violettstrablung vervendeten Versuebstieren.

In frliberen Studien iiber das Verbalten von Blutgefassen bei
versebiedenen tbermiseben Hautreizungen gestatteten die Metbo-
den keine differeuzierte Untersuebung der iieripbercn Blutgefasse
(mit diesem Begriff sind bier Kapillaren, Siuusoide und sinuose
Blutgefas.se gemeint). Eine Ausnahnie stellen die Untersuebungen
von T. Sjostraxd (1934, 1935) dar, velcbe sicb jedocb auf den
Einfluss versebiedener Lufttemperatureu, also auf generclle
Kalte- Oder AVarmereize bezieben.

‘ Der Redaktion am 14. September 1940 zugegangen.

184

GOSTA VON REIS U. ERITIOP SJOSTRAND.

Dieser Autor tmtersuchte mit derselben Methode wie wir bei
Mausen die Blutmenge in peripbercn Blutgefassen der Nebennieren,
Leber, Nieren und Skelettmuskulatur, nacbdem die Tiere I'/j — 3
Stunden in einer Kammer mit einer Lufttemperatur von entweder

+ 1° b 6° C Oder + 35° f- 37° C geweilt batten. Er erbielt

dabei eine Zunahme der peripberen Blutmenge in den Nebennieren,
der Leber und der Skelettmuskulatur bei Kalte, in der Leber und
den Nieren bei Warme.

Sonstige Untersuchungen auf diesem Gebiet sind meistens mit
Hilfe von Pletbysmograpbie, Kalorimetrie oder Tbermostromubr
vorgenommen worden. Hierbei wurden die Reaktionen in Arterien,
peripberen Blutgefassen und Venen gleicbzeitig registriert, imd
ein differ enziertes Studium der Reaktionen der einzelnen Blut-
gefasstypen var nicht moglicb.

Dass eine lokale tbermiscbe Hautreizung auf Blutgefasse ein-
vdrken kann, -welcbe von dem gereizten Hautgebiet mebr oder
weniger weit entfernt sind, zeigt schon der Versuch von Brown-
Seqxjard und Tholozan (1858) uber die sog. konsensuelle Innerva-
tion. Die Beobacbtungen dieser Eorscber sind spater durcb eine
ganze Reihe von pletbysmograpbiscben Untersuchungen (A.
Mosso 1874, .ERAN<?ors-FRANCK 1876, U. Mosso 1889, Amitin
1897, Hewlett, van Zwaluwenburg und ]\Iarshall 1911 und
Pickering 1931) sowie bei kalorimetriscben Bestimmungen von
Stewart (1911) bestatigt worden.

Was u. a. die Hautgefasse betrifft, so sollen durcb derartige
Reflexe nicbt nur symmetrisch liegende Hautpartien von einem
lokalen tbermiscben Hautreiz in Mitleidenscbaft gezogen sondern
eine allgemeinere Reaktion in diesen Gefassen' ausgelost werden
(0. Muller 1904, Pickering 1931, Freeman 1935).

Bei Kanincben fand Wertheimer (1894) nacb Kurarisierung
und Laparotomie, dass bei Abkiiblung der Haut der arterielle
Blutdruck stieg, wabrend der Druck in der Vena renalis niedriger
^vurde. Nacb Ansicbt des Autors berubt der Effekt auf einer
Kontraktion der kleineren Blutgefasse in der Niere mit daraus
folgendem gesteigertem peripherem Widerstand. Nacb Dener-
vierung der Niere ging der Blutdruck in der Arterie und Vene voll-
kommen Hand in Hand.

Mittels Tbermostromubr konstatierte Rein (1929) am Hund,
dass Kaltereizung der Haut mit Eis in der Nabe der Nasenflugel
eine Steigerung des Blutstroms in der Art. carotis comm, um 40 %
mitsichbracbte. Bei genereller Abkiiblung und wabrend zweck-

EIKFLUSS EOKALEU PliVSIKALlSCUEn U. CIIEMISCHEK UAUTREIZE. 185

niiissig gewalilter Karkose land der gleickc Autor (1931) eine Zu-
nahme der Durchstromung der iVrt. catotis comm, mit Blut von
im 28 %, in der Yena rcnalis von 13 % nnd in der Vena

mesonterica craniali? von nber 400 %, und Rein nnd Rosseer
(1929) stelltcn lost, dass die Durcliblutung von Vena porta und
Nierengefiissen nm 50 — ^200 % stieg.

RuHMA^*K (1927) land bci laparoskopischen Beobachtungep der
grosseren Darmgefiissc beim Mcnsclien, dass "Warme dieselben znr
Dilatation und Kiilte zur Kontraktion bringt. Da der Effekt rascli
eintritt meint der Autor, dass cs sick um einen Reflex liandle,
welcher von der Realction in den Hautgefassen avisgelost werden
soli.

In S])uterer Zeit bat man sicb neben diescr Reflexeimvirkung die
Idbglichkeit einer liormonal vcnnittclten Ferinvirkung gedaebt
(Laqvkur 1930. Bobxsteix 1931, Hoff 1931).

Metboaik.

Als Versucbstierc warden 400 — COO g wiegende IMcerscliweinchen ver-
wendet.

Die Bcstiinmung der Blutnienge in peripberen Blutgefiissen ^s•llrde
mit dor von T. Sjostrand (1934) angegebenen I\Ietliodo uusgefubrt,
ehenso wie i)ci unscrer friiheren Arbeit, weshalb wir auf dicse verweiseu
(G. vo.v Rf.is und F. Sjostraxo 1938). Die bLsher untersuchten Organe
.'lind die Leber, Nierenrindc und oin Skclettmuskol, der M.inasseter.
Boi der Faststcilung dor periplieren Blutmenge in der Nierenrindc wurde
nur die Blutmenge dcr tings um die Tubuli liegenden Blutgefiisse und
niebt die der Glomeruli bostimmt.

Anch .'^onst stimmt das Verfalircn mit dem bei unseren fruher be-
Kchricbenen Vcr.suclicn in Amvendung gebraehten iibercin.

Dcr gercizte Hautbezirk l)o.staud aus cinem 3 X 6 bis 4 x G cm gros-
sen Gobict quer iiber die Bauchhaut. Dieses Hautgebiet war am Tage
vor Ausfiihrung des Vcrsuclis rasiert worden. Der Hautreiz wurde
unter Pernoctonnarkosc appliziert. Die Bebandlung wurde an in
Riickonlage festgescbnalllen Tiercn vorgenommen und dauerte in der
Regcl 45 Min. Nacli der Bebandlung wurden die Ticrc durcb Abquet-
schung des Halsmarks odor durcb Dekapitation unmittelbar getotet.
Bei Versuclien mit Rcizung von denerviertcr Haut war die Nerven-
durchtrennung 10 — ^20 Tagc vor dem Versucb ausgefulixt worden.

Der Kalteroiz wurde mittcls cines Gummischlaucbs appliziert, durcb
welchen kaltes Wasser geleitct wurde.

Zur Warmereizung wurden toils ahnlicbe Verfahren mit warmem
Wasser, toils eine Warmclampe (oGIorj'*) verwendet, welcbc Warme-
strahlen von sowolil grosserer wie geringercr Wellenlange aussendet
und in verschiedenen Abstiinden von den Ticren angebraebt wurde.

13 — 'i01323. Acta ]ihvs. Scandinav. VoL 1.

186

GOSTA VON EEIS U. FKITIOP SJOSTBANB.

Abb. 1. Qaerschnitt d arch den M. masseter. Die Mnskelkapillaren verden pnnbt-
fSrmig projiziert. Schnittdicke: 20 [i. F&rbnng: Ortho-ToJidin-
wasserstoffenperoxyd.

Bei Versuchen mit der Warmelampe wurde die Warmestrahlung mittels
eines zweckmassig geformten, mit Wasser gefiillten Blechgefasses be-
grenzt. Die Lufttemperatur in der Umgebung des Tieres unter diesem
Gefass stieg dabei um ungefabr 1° C.

Bei der Warmebestrablung wurde die Hauttemperatur iniierbalb
des bebandelten Gebiets mit einem Thermoelement (Konstantan-
Kupfer) gemessen.

s Bei der Bestimmung der Anzahl offener Kapillaren pro Flachen-
einheit im Querschnitt des M.masseter wurden 20 {jl (Lcke Querschnitte
des Muskels verwendet. Nach Farbxmg der roten Blutkorperchen mit
Ortho-Tolidinwasserstoffsuperoxyd wurde die Anzahl blutgefiillter
Kapillaren durch Zahlung imter dem Mikroskop bei imgefahr 540facher
Vergrosserung bestimmt, wobei wir uns eines im Okular angebrachten
Zahlquadrats bedienten. Durch das Zahlquadrat wurde von dem
Muskelquerschnitt eine quadratische Flache von 0.082 mm® abge^enzt.
Die Kapillarenanzahl wurde in 25 derartigen Quadraten von wenigstens
10 verschiedenen Schnitten und von verschiedenen Partien des Quer-
schnitts gezahlt.

Bei derartigen Bestimmungen ist es von grosster Bedeutung, dass
die Zahlungen an guten Querschnitten vorgenommen werden, _ da
schrage Schnitte ja dazu fiihren miissen, dass die Schnittflache einer

EIKFLUSS LOKALER PHTSnfAlISCHER U. CHBMISGHBR HADTREIZE. 187

Abb. 2. SchrSgschnilt darch den M. masseter. Die Kapillaren irerden nicht punkt-
lormig projiziert, sondern an den Blntkdrperchen in der sebarfeingestellten Ebene
zeichnen sich die BlntkSrperchen in den tieferen Partien des Schnitts diffns ab.
Ein tJnterschied in der Form des Mnskelfaserquerschnitts gegennber Abb. 1 wird
trotzdesscn nicbt sichtbar. Scbnittdicke nnd Farbnng we Abb. 1.

jeden Muskelfaser grosser und damit die Kapillarenanzahl pro Flaclien-
einheit geringer livird.

So sinkt die Anzahl Kapillaren pro Flacheneinheit um 2 % bei einer
Schnittschragheit von 10°, um 15 % bei einer solcben von 30° und um
40 % bei 45°. Hieraus gebt ber\mr, dass ein geringerer Grad von
Schnittschragheit keine grossere Fehlerquelle mitsichbringt, dass aber
letztere schon bei massigen Schragheitsgraden hochst betrachtlich wird.

Um diese Fehlerquelle zu eliminieren ^vurden 20 (x dicke Schnitte
vervendet, 'vvas von T. Sjostband empfohlen worden vrar. Wenn man
namlich mit einer Kapillarweite von ungefahr 2.5 rechnet, findetman,
dass eine Schnittschragheit von 15° zur Folge hat, dass die untere
Grenzflache einer quer durch den Schnitt verlaufenden Blutkorper-
chensaule ganze zvrei Kapillardurchmesser im Verhaltnis zur oberen
Grenzflache seitlich verschoben wird. Eine solche Verschiebung lasst
sich natiirlich bei scharfer Einstellung verschiedener Tiefen im Praparat
leicht entdecken. Unsere Bestimmungen sind an Schnitten ausgefiihrt,
VO der Schnittschragheitsvinkel maximal ungefahr 15° betrug.

Eine andere Moglichkeit, den Grad der Querschneidung zu beurteilen,
bietet die Form der Schnittflachen der Muskelfasern. Aus mehreren

188

GOSTA VON REIS U, ERITIOE SJOSTRAND.

Griinden gestatten indessen diese nur sehr unsichere Schlussfolgerungen.
So treten die einzelnen Muskelfasern, welche iingefarbt sind, nicht im-
mer distinkt liervor tmd lassen sich oft nicbt mit Sicberheit unter-
scbeiden. Dadurcb dass die Querschnitte der einzelnen Muskelfasern
eine unregelmassige, abgerundet-kantige Form baben,kann die Schnitt-
scbragbeit einen erbeblicben Grad erreicben, obne dass eine deutlicbe
Formanderung eintritt. Dies gebt aus einem Vergleicb der Abb. 1 und
2 bervor. Abb. 1 stellt einen guten Querscbnitt dar, wabrend Abb. 2
einen scbraggescbnittenen Muskel zeigt, wo die Scbnittscbragbeit
durcb den Verlauf der Kapillaren deutlicb markiert wird, wo man aber
lediglicb aus den Scbnittflacben der Muskelfasern den Grad der Schrag-
beit nicbt sicber entnebmen kann.

Da man an dixnnen Scbnitten den Grad der Scbnittscbragbeit nur
durcb Beobacbtung der Scbnittflacben der Muskelfasern beurteilen
kann vermisst man daber bei diesen eine effektiv^e Kontrolle des Scbnitt-
schragbeitswinkels; bierdurch wird eine erbeblicbe Feblerquelle einge-
fiibrt, was nicbt von alien beacbtet worden zu sein scbeint, welcbe sicb
dieser Metbode bedient baben.

Aus verscbiedenen Griinden wurden bei den Bestimmungen der
Anzabl offener Kapillaren in der Muskulatur keine Metboden mit
Vergleicb mit Standardpraparaten oder Vergleicb zwiscben den ein-
zelnen Praparaten verwendet. Einserseits sind namlicb die Variatio-
nen der Blutmenge und Dicbte dieser Gefasse von einer erbeblicb ge-
ringeren Grossenordnung als es z. B. bei peripberen Blutgefassen in
der Leber und Nierenrinde der Fall ist, wesbalb die Differenzen weniger
bervortreten, andererseits bieten die Muskelpraparate ein bunt zusam-
mengewiirfeltes Bild von miteinander abwecbselnden Partien von Quer-,

, Scbrag- und Langsscbnitten, wesbalb der visuelle Eindruck bei der
fiir eine solcbe Beurteilung erforderlicben tfbersicbt des Praparats ein
hochst uneinbeitlicher wird.

Um einen gewissen Masstab fiir die Genauigkeit zu erbalten, welcbe
bei den Einzelbestimmungen obwaltet, bat diejenige Person, welcbe
durcbweg die Bestimmungen bei samtlicben bier veroffentlicbten Serien
ausgefiibrt bat, die Anzabl offener Kapillaren in zwei Muskeln bei drei
bzw. vier verscbiedenen Gelegenbeiten mit mebreren Tagen Zwiscben-
raum und obne von den Doppelbestimmungen zu wissen bestimrat.
Die Eesultate der Bestimmungen geben aus Tab. 1 bervor.

Diese Werte zeigen eine Variation zwiscben den einzelnen Bestim--
mtmgen von in beiden Fallen maximal zwiscben 6 und 7 %.

Die Anzabl offener Kapillaren pro mm* Querscbnitt wurde in 11
Muskeln bei zwei verscbiedenen Gelegenbeiten bestimmt. Die Ab-
weicbung zwiscben den beiden dabei erbaltenen Werten betrug fiir
jeden einzelnen Muskel im Jlittel 7.S % ± l.os. Stellt man die 11
Werte, welcbe bei jeder Gelegenbeit erbalten worden waren, zu zwei
Serien zusammen und vergleicbt man die IVIittelwerte dieser Serien, so
findet man eine Abweicbung zwiscben denselben von 1.8 %. Die
Abweicbungen zwiscben bei verscbiedenen Gelegenbeiten ausgefiibrten
Bestimmungen sind somit nicbt systematisch, sondern gleicben bei genii-
gend grossen Keiben einander aus.

EINFLUSS LOKALBR PHYSIKALISCHER O. CHEMISCHER HAETREIZE. 189

Tab. 1.

W’iederholie Besixmmungen der Anzahl offener Kapillaren pro mm"
Querschnitt ziueier Mjiskeln.

: Bcstimraiing Nr.

M. niiisseter

Nr. 1

Nr. 2

1 G80 ± 39

1 410 ± 41

i 2 .

1G30 + 3G

1 340 ± 53

1 580 ± 49

1 320 + 42

i 4

1 GOO + 37

Die statistisclien Berechuungen wiirden nach folgenden Formeln
vorgenominoxi;

Bci Bestimmung der Anzahl offener Kapillaren in Quersclmitten des
IM.masseter wurde die Strcming nach der Formel:

(a = Abwcichung vom Mittelwerfc; n = Anzahl Beobachtungen, hier
immer = 25, wcshalb die diesbezugliche Korrektur angebracht wurde)
berechnct.

Der mittlerc Fehler des arithmetischen Mittels wurde nach der
Formel:

(2) ■ e(M) = ±-^

bestimmt.

Bei dem Vergleich zwischen den Werten der verschiedenen Versuchs-
reihen fanden folgende Formeln Yerwendung:

(3) • • - = (A,-MJ^ + i (Bj-Mb)=]

(•i) ^ (Ma — Mb) = cr|/— + -

(5)

Ma — Mb
i7(MA-MB)

(m und n ist die Anzahl der Werte in den beiden Serien. Ai und Bj
reprasentieren den einzelnen Wert in den betreffenden Serien. t stellt
die Mittelwertsdifferenz dar, ausgedruckt in ihrem mittleren Fehler.
Siehe rStudent# 1908 und Wahlunp 1931.)

190

GOSTA VON KBIS U. FRITIOP SJOSTRAND.

4 , /s'

■O 1 $

-c

S//

x 10
9

5
7

6
S
4
Z
2
I

I — I — I — I — \ — I — I — m — I — I — I — I — I — 1 — I — I — r

A. KontroUtiere

I I

I I
I* '

■ I

V-

T — r — j — r*-! — I — — r

A '^rw^e^rahlung\

I I
1 I

T — 1 — r — 1 — T — r* — . — — i — ; — r-*-! — i — r

C V/Srmebe^trablun^

. I - i I - X - I

Q ^ ^ Si s>

Js:

U I ui- t.

: I \ I

I Leber
i Werenrinde

I I 1

Anzahl Bluikorperchen in
iausenden pro trim^ Gewebe

Die Dfeile nnter der Absziase bezeichnen die Werte bei den verwendeten
StandardprSparaten.

Diagramm 1.

Eine statistisclie Bearbeitung der bei Verwendung des Vergleicbs-
standards erbaltenen Werte mirde infolge der Schwierigkeiten, welche
duxcb die unregebnassigen Intervalle der Standardpraparate veruisacbt
werden, nicbt vorgenommen. Eine derartige Bearbeitung erscbeint
ancb im Hinblick auf die Grosse der Differenzen zwiscben den Ver-
sncbsreiben liberflussig, welche bier in den meisten Eallen vorliegen.
Stattdessen ist die Gruppierung der Primarwerte grapbiscb veranscbau-
licbt.

EIKFLUSS LOKALER PHTSIKAlISCHEB U. CHBMISCHBR HAUTREIZE. 191

Die Pfeile nnter der AbsziBse bezeichnen die TVerte bei den verwendeten
StandardprSparaten.

Diapramm 2.

Tersnchsergebnisse.

Kontrolltiere.

Gleichzeitig mit den. in fruheren Arbeiten beacbriebenen Ver-
sucben (Hautreizung mit Senfol und TJltraviolettstraMung)
\rurden laufend Kontrolltiere untersucbt. Im Laufe von reicblich
eiaem Jahr erhielten vir eln grosses Kontrolltiermaterial, welches,
obwohl die Tiere also in verschiedenen Jahreszeiten untersucbt
Worden waren und aus verschiedenen Tierzuchtereien stammten,
nicht solche Variationen hinsichtlich der peripheren Blutmenge
aufwies, dass sie bei der Beurteilung der hier erhaltenen Kesultate

192

GOSTA VON BEIS U. FBITIOF SJOSTBAND.

Tab. 2.

AnzaJil offener Eapillaren pro mw? QuerschnUt des M. masseier.

Kontroll-

tiere

"Warme-

strahlnng

Kontakt-

■warme

Ealte

Senfol

intakte Hant

1530 ± 54

1480 ± 49

1550 + 54

1560 + 50

1440 ± 55

1290 ± 40

1190 ± 26

1350 ± 50

1850 ± 64

1260 ± 40

1300 ± 46

1290 + 42

1330 ± 42

1990 ± 37

1550 + 44

1470 ± 35

1430 + 63

1150 ± 38

1340 ± 44

1660 ± 43

1230 ± 38

1410 ± 28

1390 + 57

1290 ± 39

1400 ± 30

1760 ± 60

1320 ± 39

1430 ± 49

1390 ± 51

1360 ± 37

1490 ± 36

1950 ± 55

1540 + 62

1810 ± 39

1110 ± 46

1260 + 36

1140 ± 34

1900 ± 51

1530 ± 52

1580 ± 51

1270 ± 48

1430 ± 37

1360 ± 41

1770 ± 41

1440 + 51

1440 ± 52

1420 ± 52

1400 ± 43

M = 1360

1460 ± 49

1790 ± 37

1520 ± 39

1340 ± 53

1320

1610 + 53

1510 + 61

1690 i 52

1460 ± 57

1760 ± 36

1680 ± 37

1660 + 49

1250 + 45

1480 ± 52

1190 ± 32

1360 ± 29

1220 ± 35

1660 ± 45

M = 1500

1740 + 57

1330 ± 28

1630 ± 39

1340 ± 44

1310 ± 31

1700 + 46

1580 + 36

1330 ± 28

M = 1520

1390 ± 45

1310 ± 43

1290 ± 27

M = 1330

denervierte Haut

1370 ± 41

1210 ± 53

1410 ± 47

1280 ± 40

1390 ± 54

1370 ± 40

1370 ± 44

1220 ± 46

1480 ± 50

1350 ± 45

1180 ± 47

1220 ± 31 j

1440 ± 34

1150 + 36

1420 ± 39 ;

1360 ± 38

1310 ± 42

1140 ± 45 1

1420 ± 46

1350 + 43

1190 ± 46 j

1360' ± 57

1230 + 42

1230 ± 30

1270 ± 48

M = 1260

1210 + 40

1600 ± 46

1450 + 48 1

1440 + 34

1560 + 52 j

M = 1390

M = 1330 1

KIXFLITSS LOKALKR rilYSlKAUSCIlEIl U. CIIEMISCIIER HADTUEIZE. 193

ins Ge\Yiclit fallen konnlcn. Diose Kontrolltierwcrtc bctrugen im
Mittel fiir die Leber (49 Tierc) 268 000 nncl fiir die Nicrenrinde
(44 Tiere) 363 000 Blutk6q)crchen pro nnn® Gewebe. Diagramm
1 A zeigt die Gruppicnnig dor Primarwerte.

Dio Auzahl offencr Kapillarcn pro min= Qucrschnitt im M.
masseter wirde an 19 Kontrolltioren bestimmt. Der mittlcro Wert
der Bestimmnngen botriigt 1 330. In Tab. 2 findet man die Pri-
miirwerfe.

Diesc Werte fiir Kontrollticrc liegen niebt iincrheblicb niedriger
ab die ^•on T. S-iosteaxp (1935) fiir A'ornialfciere und Tiere in
Pornoctonnarkose angegebenon Werte. was aiisschliesslicb dar-
auf berulit, das.*? vcrschicdene Prinzipicn bei der ZuhUmg dor
offenen Kapillarcn znr Anwendung kaincn. Bei Kontrollziihlung
an nnseren Prajiaratcn gemass den von T. Sjostuan'd zugrunde-
gelegten Prinzipicn rc.^^ulticrtcn Werte, wciche init denjenigen die-
ses Forschers vcillig UborcinKtimmen. Die bciden Zahlungsprin-
zipien nnfer.schoiden sich z. B. in Bezng auf Mitrechnung von
Kapillaronannstomosen, Trcnnxing diclitlicgendcr Kapillarcn und
KapiilarverzAvcignngen, Mitrcclmung einzelner Blutkorperchen,
wclehc sich niclii mit Sicherheit- auf cine quergetroffene Kapillare
ztiriickfubren lasson etc.

Da cs .cich ja nnr um relative Werte und koine absoluteii Masse
handelt, mag man die.'^on boiden Prinzipicn Berochtigung zucr-
kennen konnen, unter dcr yoraiis.sctzung, dass dicselben iihnliclie
relative Variationen an den Tag bringen. Um dies nnchzupriifeii
liabcn vir cine .Anzalil von Pnaparaten mit variierenden Blut-
rnengen naeh diesen Prinzipicn nusgeziihlt und dabci gefunden,
dns.s, olnvobl die absolutcn Werte betrachtlich alnvichen, die
prozcntuale Abwcichung zwischen den verscliicdenen Pniparaton
bei bciden Zablungsarten im grossen ganzen dicselbe war. So fan-
den wir bcksjHelswei.sc bei Bestimmnngen an funf Tieren aus je
zwei Versuch.sserion, dass die Werte dcr cinen Rcihc bei Ziihlung
nach dem liier angewendeten Prinzip im Blittel um 20.4 % holier
lagen als die der nnderen, wahrend die entsprechende Differenz
bei Ziihlung nach dem anderen Prinzip 24.5 % war.

■VVUrmorelze.

A. Wfirmostruhlung.

Bei Bcstrahlung mit der Wiirmelampc wurden zwischen 35°
und 40° C variierende Hauttemperaturen gemessen. In cinigen

194

gOsta von KEIS U. FRinOP sjSstrand.

Fallen kamen Verbrennungen zustande, wobei die Tiere kassiert
wurden. Die Serie entbalt nur Tiere obne makroskopiscb nack-
•weisbare Hautveranderungen oder solche mit leicbteren Warme-
erytbemen.

Der Mittelwert der peripheren Blutmenge betragt fiir 11 be-
strablte Tiere 350 000 Blutkorpercben pro mm® Gewebe in der
Leber und 398 000 in der Nierenrinde. Diagramm 1 C macbt die
Verteilung der Primarwerte ersicbtlicb.

Die erbaltenen Resultate zeigen, dass lokale Warmebestrablung
unter diesen Bedingungen nicht zu einer nachweisbaren Zunabme
der peripberen Blutmenge in der Leber und Nierenrinde fuhrt.
Aucb bei denjenigen Tieren, bei welcben Hautschadigungen
eingetreten waren, Hess sick kein sickerer Effekt feststellen.

Wurden die Versucksbedingungen in der Weise geandert, dass
die Zimmertemperatur von 18° C bei dem obenerwaknten Versuck
auf 25 — 30° C erkokt "wurde, dann erkielt man abweickende
Resultate. Die Mittelwerte der peripkeren Blutmenge in der
Leber und Nierenrinde wurden da fiir 10 Tiere 560 000 bzw.
842 000. Diagramm 1 B zeigt die Anordnung der Primarwerte.
Aus diesem Diagramm geht indessen hervor, dass der Mittelwert
fiir die peripkere Blutmenge in der Leber auf Grund der grossen
Streuung nickt reprasentativ ist.

Der Mittelwert der Bestimmungen der Anzahl offener Kapillaren
pro mm® Querscknitt des M. masseter wurde 1320. Die Primar-
werte findet man in Tab. 2.

Der mittlere Wert fiir die Nierenrinde ergibt eine Zunakme
mif 130 % des entspreckenden Kontrolltierwerts, wakrend sick
eine Veranderung der Kapillarenanzakl im M. masseter nickt
sicker feststellen lasst.

Bei diesen Versuchen konnte im Gegensatz zu den Versucken
bei der niedrigeren Zimmertemperatur eine Erkokung.der Rektal-
temperatur der Tiere um 1 — 2° C nachgewiesen werden.

B. Kontaktw&rme.

Eine Temperatur des durck den Gummiscklauch fHessenden
Wassers von ungefakr 50° C erwies sick als der starkste Warme-
reiz, welcken die Haut ohne Verbrennung ertragen konnte. Bei
Temperaturen von 55 — 60° C kamen regelmassig Warmescka-
digungen von derselben Art wie bei Versucken mit allzu intensiver
Wafmestraklung zustande. Bei Anwendung von 50gradigem

EINFLUSS LOKALEK PHYSIKALISCHEE U. CHEMISCHER HAUTREIZE. 195

Wasser erliielten vrii keine Steigerung der Eektaltemperatur
(Zimmertemperatur 18 — 20° C).

In einer Serie, bei ■vvelchcr Wasser mit einer Temperatur von
50° C zur Anwendung gelangte, rcsultierten folgende Mittehverte
(16 Tiere): fiir die periphere Blutmengc in der Leber 325 000 und
in der Nierenrinde 270 000 Blutkorperclien pro nim» Gevrebe.
Diagranim 2 D veranschanlicht die Gruppierung der Einzehverte.
Der mittlere Wert fiir die Anzahl offener Kapillaren im M.masseter
(9 Tiere) betrug 1 360 (sielie Tab. 2).

Die erhaltenen Werte veisen keine sicker bestimmbaren Unter-
sckiede gegcnubcr cntsprcckenden Werten von Kontrolltieren
auf.

Kiiltereize.

Zwci Tierscrien wurdcn Kiiltereizen von versckiedener Inten-
sitat ausgesetzt. Bei der eincn Serie vrurde Wasser mit einer
Temperatur von -f- 5 f- 10° C durch den Gummiscklauck geleitet.

Die Mittclvrerte der Bliitmenge in pcripkeren Blutgefassen be-
tragen bei diescr Serie (18 Tiere) fiir die Leber 652 000 und fiir die
Nierenrinde 323 000 Blutkorperchen pro mm® Gewebe (sieke
Diagramm 2 E). Die Anzakl offener Kapillaren pro mm® Masseter-
qucrscknitt ist im Mittel (15 Tiere) 1 710 (sieke Tab. 2).

Dicse Werte zeigen cine Zunahme der peripkercn Blutmenge
in der Leber mit 150 % dcs entspreckenden Mittelverts bei den
Kontrolltieren, wakrcnd sick ein sickerer Einfluss auf die Nieren-
rinde nickt nackweiscn lasst. Im M.masseter ist die Anzakl offener
Kapillaren im Vergleick zu den Kontrolltieren um 29 % gestiegen.
Die Differenz der mittleren Eekler (t) ziviscken dieser Serie und
der Kontrolltierserie betriigt 8.6, weskalb die Differenz der Mittel-
vrerte statistisck gcsickert ist.

Um den Einfluss einer direkten Abkiihlung des Leberparen-
ckyms in moglichst grossem Ausmass zu elimmieren mirde der
Gummiscklauch bei drei Tieren so weit kaudalvrarts "wde nur mog-
lick angebrackt. Hierbei erliielten vrir folgende Werte fiir die
peripkere Blutmenge: in der Leber 850 000 fiir samtlicke Tiere
und in der Nierenrinde 620 000 bzv. 430 000 und 430 000 Blut-
korperchen pro mm® Gewebe.

Es wurde also auck bei diesen Versucken ein iiknlicker Einfluss
auf die peripkeren Blutgefiisse der Leber konstat’ert. Die peri-
pkere Blutmenge in der Nierenrinde vreist eine eventuelle Ver-
mehrung auf.

196

GOSTA VON BEIS U. FRITIOF SJOSTRAND.

Die Tiere der anderen Serie •wTirden einem niilderen Kaltereiz
ausgesetzt, indem Wasser mit einer . Tempera tur von 15° C
durcli den Gummisclilauch geleitet "wurde. Hierbei erhielten vir
folgende llittelwerte (9 Tiere): in der Leber 556 000 nnd in der
Nierenriade 312 000 Blutkorperchen pro mm® Gewebe (siehe
Diagramm 2 F). In dieser Serie tritfc also ein etwas niedrigerer
Mittelwert der peripberen Blutmenge in der Leber anf als bei
dem starkeren Kaltereiz, aber es -vnirde gleicbwohl eine Ver-
doppelung gegeniiber dem JEttelwert der Kontrolltiere erreicht.
Die peripberen Blutgefasse der Nierenrinde lassen aucb in diesem
Yersucb keinen sicberen Einfluss erkennen.

Kaltereizung’ denerviei’ter Haut.

Um ebenso vie bei den friiher veroffentlicbten Versucben die
Bedeutung der Hautinnervation fiir die Ubertragung des Eeizes
ausfindig zii macben vTirden Experimente iiber den Effekt von
lokaler Kaltereizung denervierter Haut angestellt.

Hierbei murde Wasser mit einer Temperatur von -j- d f- 10° ^

verwendet, -welches durcb den Gummiscblauch geleitet wurde.
Die blittelwerte der Bestimmungen der peripberen Blutmenge
an 12 Tieren waren: Leber 318 000 und Kierenrinde 271 000
Blutkorperchen pro mm® Gewebe (siebe Diagramm 2 G). Die
Anzabl offener Kapillaren pro mm® Querschnitt des M.masseter
betrug im Mittel 1 390 (siehe Tab. 2).

Diese Werte stimmen ja mit den eutsprecbenden Kontrolltier-
werten gut iiberein. Die Differenz der mittleren Eebler (t) zwischen
dieser Serie und der Versuchsreihe mit Kaltereizung von Haut
mit intakter Innervation betragt fiir die Muskel werte 5.9, wes-
balb die Differenz der blittelwerte statistiscb sicber ist. Der
EffeM auf 'peri'pTiere Blutgefasse in der Leber und im M.masseter
scheint also bei loJcalen Kdltereizen von einer intaJcten Hautinnerva-
tion abhangig zu sein.

Einfluss lokaler Hautreize mit Senfol und TJlti’aviolett-
stralilung auf die Anzahl ofiTener Kapillaren
pro mm® Querschnitt im IM. masseter.

A. Intakte Hautinnervation.

Folgende Besultate mogen als Erganzung der friiher veroffent-
licbten Versuche iiber den Einfluss lokaler Hautreizung mit Senfol

EINJ?LUSS LOKALER PHYSIKALISCIIBR 17. CHEMISCHER HAUTREIZE. 197

tuid Ultraviolettetralilung aiif die periphere Blutverteilung in der
Leber und Nierenrinde (G. VOK Eeis und F. Sjostrand 1938)
bier angefiihrt warden:

Bei lokaler Hantreizuug niit Senfol wabrend 20 — 45 Min. wurde
der Mittehvert der Bestimmungen an 12 Tieren 1 500 Kapillaren
pro nim= Querscbnitt des M. niasseter (siebe Tab. 2). Die Diffe-
renz der niittlereii Febler bei dieser Serie nnd der Kontrolltierserie
betragt 3.2. wesbalb die Differenz der Mittelwerte statistiscli
gesicbert ist. Die Zunabme der Anzabl offener Kapillaren macbt
iin Mittel 13 % aus.

15 Tiere, welcbe unmittelbar nacb 45 blin. dauernder Bestrab-
lung mit ultra\'ioletteni Licbt getotet worden waren, batten im
Mittel 1 520 offene Kapillaren pro mm- Masseterqnerscbnitt
(siebe Tab. 2). Die Zunabme betragt im Vergleicb zu den ent-
sprecbenden Kontrolltierwerten bier im blittel 14 %, und da die
Differenz der mittleren Febler bei diesen beiden Serien 4.4 aus-
macht, ist die Vermebrung statistiscb sicber.

Eeibt man diese Eesultate den bei friiberen Versucben iiber die
peripbere Blutmenge in der Leber und Kierenrinde bei denselben
Zustanden erhaltenen an, so gelit bervor, dass die peripberen
Blutgefiisse sicb bei lokaler Hautreizung mit Senfol und Ultra -
violettstrablung sowolil in der Leber, wie in der Nierenrinde und
im M.masseter dilatieren.

B. Denervierte Haut.

Bei entsprecbender Reizung in Hautbezirken, vrelcbe ibrer
segmentalen Innervation beraubt worden waren, erbielten vir
mit Senfol (8 Tiere) im Mittel 1 260 und bei Bestrablung mit
ultraviolettem Licbt (11 Tiere) 1 330 offene Kapillaren pro mm-
Masseterquerscbnitt (siebe Tab. 2). Die Differenzen der mittleren
Febler bei diesen Serien und den entsprecbenden Versucbsreiben
von lokaler Hautreizung mit Senfol und Ultra violettstrablung
an Haut mit intakter segmentaler Innervation betragen 3,4 bzv.
3,2, wesbalb die Differenzen zwiscben den Serien statistiscb ge-
sicbert sind.

Diese Werte zeigen, dass man bei Keizung von denervierter Haut
keinen sicberen Einfluss auf die Kapillaren im M.masseter erbalt,
wesbalb die obenbescbriebenen Effekte auf diese Blutgefasse
dutch die Hautnerven vermittelt werden diirften.

198

QOSTA VON NEIS U. ERITIOF SJOSTEAND.

Erorterung der Versuchsergebnisse.

Durch. die Ausfiilirung der hier bescliriebenen Versuolie in
Narkose 'wird ausgescHossen, dass die erzielten Effekte .eine Folge
von psycMsclier Exzitation durch die starke Reizung der Sinnes-
nervenendigungen ■waxen. Als Narkosemittel wurde Pernocton ver-
wendet, welches an sich die periphere Blutverteilung nicht nen-
nenswert verandert (T. Sjostband 1935, Lindgken 1935), wes-
halb -wir bei den Versuchen von einem Zustand in den peripheren
Blutgefassen ausgegangen sind, welcher dem bei nichtbetaubten
Tieren recht gut entsprechen diirfte.

Der friiher vorgenommenen Kontrolle nach (R. von Reis und
F, Sjostkand 1938) bringt die Dener-vderung keine Zirkulations-
storung in der Haut mit sich, welche beispielsweise fiir die Ent-
stehung, Freimachung oder Resorption eventueller reizvermit-
telnder Substanzen Bedeutung haben konnte, weshalb die Ver-
suche demonstrieren diirften, dass der Effekt von lokaler Haut-
reizung mittels Kalte auf periphere Blutgefasse in der Leber und
dem M. masseter des Meerschweinchens durch das Nervensystem
vermittelt ■wird,

Es ist wohl wahrscheinlich, dass der bei lokaler Hautreizung
mit Senfol, Ultra^vdolettstrahlung und Kalte wahxnehmbare Effekt
auf die KapHlaren im M.masseter sich nicht auf diesen Muskel
beschrankt. Der M. masseter entspricht ja nicht denjenigen Haut-
segmenten, in welchen der Hautreiz appUziert worden war, und
nimmt offenbar keine Sonderstellung den ■iibrigen Skelettmuskeln
gegeniiber ein. Es erscheint uns somit berechtigt, die Resultate
der Bestimmungen der Anzahl offener Kapillaren in diesem Muskel
in gewissem Masse auch auf die iibrige Skelettmuskulatur zu iiber-
tragen.

Die tibereinstimmung zwischen unseren Resultaten bei lokaler
Hautreizung mit Kalte an Meerschweinchen und den von T. Sjo-
STEAND an Mausen bei genereller Kaltereizimg erhobenen Befun-
den ist augenfalhg. So fand dieser Autor ebenfalls eine Zunahme
der peripheren Blutmenge in der Leber, wahrend sich ein sicherer
Effekt auf die Nierenrinde nicht konstatieren liess. Diese Uber-
einstimmung zeigt offenbar, dass man bei Kaltereizung in einem
kleineren Hautbezirk Reaktionen in den peripheren Blutgefassen
von entsjirechender Art wie bei einem generellen Kaltereiz aus-
losen kann. Hinsichtlich Warme durften T. Sjosteands Befunde
am ehesten mit den hier beschriebenen Versuchen mit lokaler

EINFLUSS LOEALER PHYSIKALISCIIER U. CHEMISCHER HAETRBIZE. 199

Warmereizung bei hoberer Zimraertemperatur zu vergleichen sein,
bei welcben eine. Erbohung der Korpertemperatur erzielt wurde.
Bei eincm solcben Vergleicb liegt eine gewisse tibereinstimmung
der Resultate vor.

Die bei der \'orliegendeii iVibeit erbaltenen Resultate steben
alletdings nicbt in Einldang zii deni, was man friiberen Versucben
mit Pletbysmograpbie nach erwarteu sollte. Die Erklarung dieser
Verscbiedenbeit kann teils darin zu sucben sein, dass man bei den
erwabnten plotbysmograpbiscben Dntersucbungen vor allem die
Reaktionen der Hautgefiisse registriert batte, welcbe ja bin-
sicbtlicb ibrer speziellen Eunktion im Zusammenbang mit der
Tbermoregulation eine Sonderstellung einnebmen diirften, und
teils darin, dass zu- und ableitende Blutgefasse ganz verscbieden
von den peripberen Blutgefassen reagieren, und dass die Reaktio-
nen der ersteren Gefasse die der letzteren bei der Pletbysmo-
graphie verdecken. Letzteres erscbeint recbt plausibel, da es
bei unseren Kaltereizungsversucben auffiel, wic extrem blass die
Organe makroskopiscb aussahen, was einer geringen Blutfiillung
der zu- und ableitenden Blutgefasse zugescbrieben werden muss.
Dieser Umstand bildet auch ein Beispiel dafiir, irie unsicber es
sein kann, aus der Farbe eines Organs auf die periphere Blut-
menge desselben zu scbliessen.

Man kann sicli vorstellen, dass eine gesteigerte Durchstromung
mit Blut durcb zu- oder ableitende Blutgefasse durcb eine rascbere
Zirkulation in den peripberen Blutgefassen obne eine Zunabme der
peripberen Blutmenge zustandekame, und zwar infolge von Reak-
tionen in zu- und ableitenden Blutgefassen und daraus folgenden
veranderten Druckverbaltnissen. So kann man sicb eine vermebrte
Durcbstromung lediglicb als Folge einer Dilatation von Arterioli
denken. Desbalb lessen sicb bei der Registrierung raittels Tber-
mostromubr keine sicheren Schlussfolgerungen fiber die periphere
Blutverteilung ziehen.

Ein Gegenstuck in den peripberen Blutgefassen zu der Steige-
rung der Durcbstromung in der Vena renalis, welcbe Rein, Rein
und Rossler bei genereller Abkuhlung konstatiert baben, konn-
ten wir bei lokaler Kaltereizung und aucb T. Sjostrand bei gene-
reller Abkuhlung nicbt nacbweisen, was, von dem oben angefiibr-
ten abgesehen, darauf zuriickzufuhren sein konnte, dass man
sicb einen vermelirten Blutstrom durcb die Nieren z. B. durcb
die Glomeruli vorstellen kann, obne eine entsprecbende Dila-
tation der peripberen Blutgefasse rings um die Tubuli.

200

GOSTA VON REIS U. FRITIOE SJOSTRAND.

Der von Rein, Rein und Rossler konstatierten erhohten
Durchblutung durch die Vena porta sckeint dagegen eine Zunahme
der periplieren Blutmenge in der Leber zu entsprechen, welche
man sowolil bei genereller (T. Sjostrand 1935), wie den vorlie-
genden Versucben nach bei lokaler Kaltereizung erbalt.

Es ist auffallend, dass bei Hautreizen verscbiedener Art die
peripberen Blutgefasse in den bier studierten Organen in bobem
Grade verscbieden beeinflusst werden. So ziebt Hantreizung mit
Senfol und Ultraviolettstrablung eine Zunabme der peripberen
Blutmenge in sowobl der Leber, wie in der Nierenrinde und Musku-
latur nacb sicb, wahrend ein Kaltereiz auf dieselben Blutgefasse
nur in der Leber und Muskulatur einwirkt. Auch starke Warme-
reize geben keinen nacbweisbaren Effekt, ausser im Zusammen-
bang mit Hypertbermie, und dann bauptsacblicb in der Nieren-
rinde. Im letzten Eall ist nicbt untersucbt worden, ob der Effekt
von intakter Hautinnervation abbangt. Diese Verbaltnisse scbei-
nen zu der Annabme zu berecbtigen, dass die Wirkungen von
Hautreizen mit verschiedenen Reizmitteln in verscbiedener Weise
innerbalb des Nervensystems vermittelt werden konnen.

Zusammenfassimg.

Die periphere Blutmenge in der Leber und Hierenrinde sowie
die Anzabl offener Kapillaren pro mm® Querschnitt des M.masse-
ter vmrde an Meerscbweinchen nacb lokaler tbermiscber Haut-
reizung in Pernoctonnarkose bestimmt. Eriiber veroffentbchte
Bestimmungen der peripberen Blutverteilung in der Leber und
Nierenrinde nacb Einwirkung von lokaler Hantreizung mit Senfol
und Ultraviolettstrablung wurden durcb Bestimmungen der An-
zabl offener Kapillaren pro mm® Masseterquerscbnitt erganzt.

1. Lokale Bestrahlung mit einer Warmelampe bracbte keinen
nacbweisbaren Effekt auf die periphere Blutverteilung in der Leber
und Kierenrinde mit sicb, ausser im Zusammenbang mit deni
Eintreten von Hypertbermie (1 — 2° C), wo sicb eine Steigerung
der peripberen Blutmenge in der Kierenrinde mit 130 % der
entsprecbenden Kontrolltierwerte nacbweisen liess. Ausserdem
lag ein infolge von Streuung der Werte nicbt sicherer Einfluss
auf die Leber vor, wahrend kein Effekt auf die peripberen Blut-
gefasse im M.masseter festgestellt werden konnte.

2. Lokale "Warmereizimg -mittels Kontaktwarme (obne Steige-

EISFLb'?S LOKALER PliySIKALISGllER U. ClIEMISCHER HAUTREIZE. 201

rung der Kektaltoniperature) hatte keine sichere Beeinflussung
dcr periphercn Blutgcfasse. in der Leber iind Nierenrinde zur Folge.

3. Bci lokalcn Kaltereizen stieg die pcripbere Blutraenge
in der Leber mit 150 % de^ entspreclienden Kontrolltierwerte,
wahrend sick ciji Einflu$s auf die Nicrenrinde nicht beobackten
]ie$s. Die Anzalil oifener Kapillaren im M. masseter nakm um
29 % zu.

4. Lokale Kfilfcrcizung von denervicrter Haut brachte keinen
sichercn Einflups aul peri])liere Blutgefiisse in der Leber, Nieren-
rinde imd im M.ma?5eter mit sick, woraiis gefolgert ^vi^d, dass der
Kffekt nuf die Leber und Muskulatur bei Kaltereizung der Haut
durck die Ilautncrven vermittelt wird.

5. Die Anzahl offoncr Kapillaren im kl.masscter stieg bei:
a) lokalor Haiitroizung mit Senfol um 13 %, und

h) lokalor Bostraklung mit ultraviolettem Lickt um 14 %.

G. Dicpc Effckte blieben aus, wenn die Hautreizung mit Senfol
odor Ultravioletfstrahlung auf Haut nppliziert wurde, welcke ikrer
segmcntalcn Innervation beraubt worden war.

Schrifttum.

Amitin, S., Z. Biol., ISO", 35, 13.

Bornstetn-, a., Z. KrcislForPch., 1931, 23, 120.

Brown-^equaro und Thoix>zan, J. Pliysiol. dc Bro^vn-Sequa^d, 1858,
407.

pRAyi^ois-FitAXCK, Travamv du Laboratoire de Marcy, 1876, 2, 1.
Fi'.EE.MAy, K. E., Aincr. .T. Phy.siol., 1035. 113, 384.

Hewlett, A. IV., J. G. van Zwaluwenburg und M. Marshall,
Arch, intern, Med., 1911, S, 591.

Hoff. F,, ?Iuricii. mod. Wschr., 1931, 78, 314 und 350.

Laqueuk, a., Fortschr. d. Thcrap., 1930, 6, 309.

Lindguen, a., Acta chit. Scand., 1935, 77, Suppl. 39.

Mosso, A., Von cinigen neuen Eigensekaften dcr Gefasswand, Leipzig
1874, zit. nach Francois-Frank, 1876.

Mosso, U., .tVreh. it.al. Biol., 1889, 39.

Muller, 0., Dtsch. .-Irek. klin. Med., 1904 — 05, 82, 547.

Pickering, G. W., Heart, 1931, 16, 115.

Been, H., Z. Biol., 1929, 89, 319.

— , Ergebn. Physiol., 1931, 32, 28.

— , und R. Rossleb, Klin. Wsebr., 1929, 8, 1457.
v. Rf-is, G., B. P. SiLFVERSKioLD, F. Sjostrand und T. Sjostband,
Skand. .Vreh. Physiol., 1938, 79, 134.

V. Reis, 6. und F. Sjostrand, Ebenda, 1937, 77, 71.

— , — , Ebenda, 1938, 79, 139.

14—401323. Acta phi/s. iycandinav. Yol.l.

202

GOSTA YON EEIB U. FRITIOF SJOSTRAND.

RtTHMANN, W., Z. ges. exp. Med., 1927, 57 , 768.

Sjostrand, T., Skand. Arch. Physiol., 1934, 68 , 160.

— , Ebenda, 1935, 71 , 85.

— , Ebenda, 1935, 71 , Suppl. 5.

Stewart, G. N., Heart, 1911 — 12, 3 , 33 und 76.
Wertheimer, E., Areh. Physiol, norm. path.. 1894, 26 , 308.

From tbe Physiological Department of the Karolinska Inst., Stockholm.

Tension Changes during Tetanus in Maniniiilian
and Arian Muscle.^

U. S. von EULER and ROY L. SWANK.*

(With 13 figures in the text.)

The tension developed by a mammalian voluntary muscle
during a tetanus elicited by indirect stimulation of a moderate
frequency usually becomes maximum quicUy, and after a “pla-
teau” of varying duration gradually diminishes. In a rested white
muscle this tension may remain approximately maximum for some
10 seconds, but the tension developed by a second tetanus which
follows the first after a short interval rapidly declines because of
fatigue, although it may yield the same or even higher initial ten-
sion.

Under similar experimental conditions with the tibialis muscle
stimulated indirectly by a maximal tetanizing current at a fre-
quency of about 45 p.s., a somewhat different tension curve may
be observed. This is characterized by the usual quick rise followed
by a momentary decrease in tension and then a secondary rise,
which may attain a considerable height.

Some of the features of this phenomenon and the conditions
governing its appearance have been investigated and will be
described in the present paper.

Methods.

Cats and pigeons have been used as esq)erimental animals.
In the majority of cases the animals were decerebrated while

’ Received 2 September 1940.

* Commonwealth Fund Fellow.

15 — W1323. Acfa phys. Scandinav. Vol, I.

204

U. S. YON BOLER AND ROY L. SWANK.

anesthetized with ether. The tibialis anticus (white muscle) or
soleus (red muscle) of the cat, were prepared and its tendon con-
nected to a spring myograph. The leg was firmly fixed in a
Brown-Schuster myograph stand. The muscle tension was re-
corded on a smoked drum or on photographic paper.

The muscle was stimulated indirectly through shielded silver
electrodes on the sciatic nerve (see diagram, fig, 1) by means of
neon lamp stimulators, acting on two pairs of electrodes, one
pair for single shocks, and the other (central) pair for tetanic
stimulation. All stimuli were just supramaximal. The stimulating
cathode was placed nearest the muscle.

Fig. 1. Stimulating electrodes, a from above, b from tbe side. The nerve is slipped
on the silver electrodes, emerging from the floor of the ebony shield.

Precautions were taken to keep the animal warm and the muscle
warm and moist.

In the experiments on denervated muscles the sciatic nerve
was cut 5 to 16 days previously and the muscle stimulated directly
by silver pin electrodes inserted about 3 cm. apart in the belly of
the muscle.

Action potentials were led off from the belly and tendon of
the muscle by silver pins and recorded by means of an Abramson
torsion band oscillograph (Weber, 1939) fed from a condenser
coupled amplifier of the type described by Matthews (1938).
The oscillograph and myograph light was derived from a 4
volt lamp and accumulator and the whole recording system was
contained in a case 50 X 30 X 30 cm. (see diagram).

In a few cases action potentials w^ere recorded with a cathode
ray oscillograph.

Close arterial injections were made according to the technique
described by Brown (1938).

Time was recorded in ^/bq th second with a stronger mark at
every ^/loth second.

TEKSION CHANGES DURING TETANUS.

205

Exporlniental Results.

I. Cats.

A- Tibialis Anticus Muscle.

The appearance of the secondary rise in tetanus tension.

The effect to be described we have termed, for the sake of brevity
“secondary rise”.

A t 53 )ical secondary rise in tetanus tension in the tibialis anticus
muscle is illustrated in fig. 3. The character of the rise varied
in details, especially concerning its height, which was sometimes
hardly perceptible, and at other times showed an increase in tension
of 0.8 kg. (Fig. 5 a.) Variations also occurred with regard to the
initial dip, which was well-marked, small or absent. The appear-
ance of the »dip)> was to some extent accentuated by tbe inevitable
over-throw due to the inertia of the lever in the mechanical re-
cording system, but this did not account for the whole effect.

The increase in tetanus tension usually appeared after a few
seconds of stimulation and lasted for some 10 sec., but had in
some cases a longer duration, although it seldom exceeded 15 sec.

The most typical examples of secondary rise occurred, as a rule,
one hour or longer after the animal had been prepared. V'e also
had the impression that the effect was more pronoimced if the

206

U. S. VON BULER AND ROY L. SWANK,

Fig. 3. Cat, chloralose anesthesia, tibialis anticus. Indirect maximal single shocks,
3 sec. apart. Maximal tetanus at 30 per sec. for 1 min:

Fig. 4. Cat, decerebrated, tibialis anticus. Maximal motor nerve single shocto
2.9 sec. apart. Two 10 sec. tetani at 45 per sec. interval 1 min. First tetanus 9 min

after the preceding one.

single (maximal) twitcli tension was low, as it usually was in
preparations used for several hours.

Influence of stimulating frequency and recovery period.

It soon became evident that the phenomenon of the secondary
rise could be most regularly demonstrated when a stimulating
frequency of about 40 — 50 p.s, was used. At both lower and higher
frequencies the effect was less conspicuous, and with frequencies
below 25 p.s. and above 70 p.s. it was usually absent..

TENSION CHANGES DURING TETANUS.

207

Fig. 6. Cat, decerebrated, tibialis antious. Maximal motor nerve single tiritches
2.8 see. apart. A. motor nerve tetanus 45 per sec. for 10 sec. Interval between shocks
daring (a) 1 min. B. series of 18 tetani at 45 per sec. 0.6 duration, 1 sec, interval.
C. tetani at 45 per sec., 0.4 sec. duration, 3 sec. interval. Before and after the tetani

single shocks.

It was also found that the interval between tetani was of great
importance. If the time elapsing between two tetani was less
than one minute the secondary rise was greatly diminished or ab-
sent during the second tetanus. As this interval was increased the
secondary rise became more pronounced and reached a maximum
when the interval was 7 to 10 minutes or longer. The failure of a
second tetanus after a very short interval to show the effect nlay
be explained by the fact that the initial tension of the second
tetanus was already maximum or nearly so (fig. 4).

Action of interrufted tetani.

The development of the secondary increase in tension was
clearly demonstrated by using intermittent short tetani. Fig. 5
B shows the gradual increase in tension which appeared when the
muscle was stinaulated indirectly 18 times at 2 sec. intervals with
tetani of Yz' — 1 sec. duration. The second short tetanus shows a

208

U. S. VON EULER AND ROY L. SWANK.

decrease’ in tension as compared with, the first one, but a definite
and gradual increase in tension to maximum takes place in the
following tetani which occupy a period of 25 seconds. Following
this increase the tension falls in subsequent tetani.

If the intervals between tetani of 1 second duration is increased
to 10 seconds the secondary rise is barely detectable, but at 3 sec.
intervals the initial decrease as well as the following increase in
tension is well established even when tetani of ’•/g sec. are used (fig.
6 c). Shorter intervals between tetani do not appreciably alter
the form of the usual secondary rise curve (fig. 6).

Kg. 6. Cat, decerebrated, tibialis anticus. Sciatic nerve out 6 days previously.
Direct stimulation 60 per sec. First tetanus interrupted for 0.5 sec. every 3 sec.
Second tetanus 20 sec. after the first one.

Effect on denervated muscle.

In order to determine if the phenomenon was dependent for
its appearance upon an intact neuromuscular connection, we have
directly stimulated muscles previously denervated. In four such
experiments the secondary rise phenomenon was clearly present
(fig. 6).

Relation between the secondary rise and the single twitch tension after

a tetanus.

It was regularly observed that the appearance of the secondary
rise was closely related to the post-tetanic potentiation as described

TENSIOK CHANGES DHUINQ TETANUS.

209

by Brown nx\d Euler 1938, and others. Thus a post-tetanic
potentiation was always observed in those cases where a secondary
rise could be produced whereas a secondary rise in muscle tension
was regularly absent when no post-tetanic potentiation appeared
as a result of a preceding tetanus. The functional relationship
between these two phenomena is further stressed by the fact that
a secondary' rise could not be produced effectively in a second
tetanus unless the post-tetanic potentiation had disappeared
or greatly decreased. Usually the p. t. p. was present for 5 — 10
minutes after a 10 second tetanus. The muscle tension elicited
by single slightly supramaximal stimuli at regular intervals there-
fore served as an accurate indication of the degree of recovery of
a muscle from tetanus (fig. 5 A).

Sometimes the first few single tnutches whicli followed a tetanus
with a secondary rise were smaller than the ones wliich followed
(fig. 2), thus causing a delay in the development of the post-tetanic
potentiation. On the other hand this delay in the appearance
of the p.t.p. has been observed to some degree during a second
tctami-s when a .secondary rise was absent, and it has also been
ob.scrvcd after potassium injections, and after curare.

Action of 'potassium injections.

The relationship of the secondary rise phenomenon and the
post-tetanic potentiation led us to test the influence of potassium
upon it. Potassium chloride was administered intravenously in
isotonic solution and by way of close arterial injections. In our
experiments this procedure decreased or abolished the secondary
rise. The doses used were 1 to 6 mg. KCl, which produced a
twitch and a subsequent increase in the single muscle twitch
tension for some time, as described by Brown (1937). The effect
of single small doses of KCI intraarterially upon the initial tetanus
tension was negligible, whereas repeated doses or big single doses
produced a lasting decrease in tension.

Action of curare.

In several experiments the secondary rise phenomenon was
studied in muscles that were gradually being curarized. When the
curare action had developed to such a degree that the single
twitch tension was definitely reduced the secondary rise during
the tetanus was also reduced. Although the initial tetanus tension

210

U. S, VON EULER AND ROY L. SWANK.

was fairly well maintained there was only a small secondary
increase in tension or none at all during the tetanus and a second
tetanus, at a later stage of the curare action, showed no sign of a
secondary increase in tension (cf. Feng et al. 1938).

B. Eicperiments on the Soleus Muscle.

Since the tibiaKs anticus muscle of the cat belongs to the group
of “white” muscles, which develop a quick contraction, we thought
it desirable to investigate- the secondary rise phenomenon on a
“red” muscle. For this purpose the soleus muscle was used.

The behaviour of this muscle in a tetanus differs in several
respects from that of tibialis anticus. Thus the tetanic tension is
usually high compared with the single twitch tension; a smooth
tetanus is obtained with a tetanizing frequency of less than
16 p.s.; the tension is sustained for a much longer time; and the
same type of tension curve may be repeated after a short interval.

The same frequency of stimulation which produced a secondary
rise in the tibialis was ineffective in this respect on the soleus.
When the stimulating frequency was lowered to 16 — 26 per second
a slight rise in tension occurred during the tetanus, but it was
much less marked than that observed in the tibialis (fig. 7). In
fact we could never be certain that the response of the two muscles
was of the same type. Contrary to observations on the tibialis
two or more tetani in rapid succession produced almost identical
responses in soleus, thus giving us another indication that the
action of these muscles is different. In the tibialis the secondary

A B iKfl

Kg. 7. Cat, decerebrated, soleus. A. Two maximal motor nerve tetani at 18 per sec.
for 10 sec. 17 sec. apart, maximal single shocks 3 sec. apart. B. Tetanus 15 per sec.
followed by series of tetani of 1 sec. duration and 1 sec. apart.

TENSION CHANGES DURING TETANUS.

211

rise was accompanied by a post-tetanic potentiation of the sub-
sequent single twitches, and a series of short tetani in rapid suc-
cession showed an increase in tension. Since neither of these ac-
tions occurred as a result of low frequency stimulation of the soleus
we are inclined to believe that the small tension rise in this muscle
under the condil ions mentioned is different in character from that
in libialis.

C. Action Potentials.

Muscular action potentials have also been recorded both alone,
and simultaneously with ihc mechanical records during the de-
velopment of the secondary rise. Because of the finding of Brown
(1937) and Brown and Euler (1938) that an increase in muscle
tension after an injection of potassium or following a short tetanus
may be associated with a decrease in action potential this seemed
especially desirable. The similarity of the beha^^our of the action
potentials in these two instances was one of the reasons for assu-
ming that they were, in principle, due to a similar kind of change
in the muscle, c. g., a shift in the potassium ions of the muscle.
In the present experiments wc have found a more or less pro-
nounced decrease in spike height dtiring the increased tension con-
stituting the secondary rise, fig. 8 illustrates this point. We have
also confirmed the observation of BRO^VN and Euler that the
increased muscle tension in a maximal single tvitch following a
tetanus is regularly associated with a decrease in the action po-
tentials of the whole muscle.

During a tetanus the action potentials usually decreased gra-
dually as the muscle tension increased, but in a few cases in which
the secondary rise was weak or absent the muscle action potentials
failed to decrease. As a rule, however, the action potentials fell
slowly at first and sharply during the final fall in muscular tension
which followed the secondary rise. We have never observed an
increase in a p. during the secondary rise if the stimuli were
maximum.

During a series of short tetani there is a gradual increase in the
maximum tension developed by each tetanus so that a secondary
rise in tension results. During this increase in tension the action
potentials gradually diminish, and later when the muscle tension
starts to fall the action potentials decrease even more rapidly.
Finally when the muscle tension has fallen back to the initial value
(e in fig. 9) the a, p. may be decreased by some 30 %. In (f) the

lO to

^ 5=
■§"§
2 §
c 2

ei ^
o ^

•id
^ «4-«

O O
nS
rrt ®

c ®

S 'O

■S s

•

2 o
§>2
>-iO

o S
> «

a-f

S “
s.e

^ c
o a
1-^ 'to
o
. as

a o

g '

• <

c 2 !

•2 i.i

.* Cm «4

'3 *“ a
'“S'*

Is o i

■*2 53 '

g^J

■g-g '
c^i i

a o '

O 03
> pid

(•« o

s o
o -C

t- ®

o o

^ eo

. o

03 >

O ^

cd o
•*»

00 O

=12

as A

•■s a

-B g

CJ ®

O M

Q**-t

to u

.-, o

P«

r4<

Fig. 9. Cati docorcbratcd, tibialis antious. Upper curve action potentials, bolly-tondon lead. Lower curve myogram. 37 maximal
indirect totani at 44 per sec. for about O. l sec. each with 0.4_seo. intervals. Totani nr 1, 2, 14, 24 and 37 are shown. Single maximal

motor nerve shooks before and 1 sec. after the series of tetani.

TENSION CHANGES DUEINQ TETANUS.

213

» t ( t

^ r*' \ rnmm ^

! ! i I ! ■ f

Pig. 10. Cnt, decerebrated, soleus. Action potentials before and after, and at the
beginning and end of a 5.5 sec. tetanus at 17 per sec. Time 1/5 sec. (cathode-ray

oscillograph).

muscle tension and a. p. liave fallen even more, and the single muscu-
lar twitch about 3 secs, after the last tetanus was characteristically
potentiated and the a. p, reduced. From the curves it is obvious
that the first contraction or the first and second ones in each
tetanus bears a close relation to the maximum tetanic tension
(except in the last tetanus, f.) but do not correspond to the first
action potential. However, the first contraction of the last tetanus
and the following single twitch are in close agreement with regard
to muscle tension and action potential. It seems that the altera-
tions in the muscle which lead to the subsequent appearance of a
post-tetanic potentiation develop in a way which can be followed
by obsendng the separate action potentials, and the first contrac-
tion of each tetanus. The curves b and c in fig. 9 also make clear
that the frequently occurring initial fall in muscular tension during
a series of tetani is due to a decrease in the first part of the second
tetanus curve. As soon as the first contraction of the tetanus is
increased the maximal tension also increases and the a. p. di-
minishes. The time necessary for the secondary rise to appear is
approximately the same as that necessary for the post-tetanic
potentiation to develop, i. e, 1 — 2 sec.

In the soleus, however, we have not observed any definite
changes in the action potential during tetani of 5 — 10 sec. duration
at frequencies 15 — 25 per sec. though there was the usual rise in
tension. The action potential accompanying the first single twitch
about 1 sec, after the tetanus was mostly imchanged, (fig. 10),

214

U. S. VON EULER AND ROY L. SWANK.

and so was the tension of the twitch. We take this as an indication
that the gradual rise of tension in the soleus during a low frequency
tetanus is of another kind than in the tibialis.

II. Tension Changes during Tetanus in the Tibialis Anticus

Muscle of the Pigeon.

The secondary rise in tension which developed during tetanus
in the tibialis anticus muscle of the cat was also observed in the
corresponding muscle of the pigeon. The response occurred reg-
ularly in this muscle of decerebrated pigeons when tetanizing
frequencies of the same order as those used in the experiments on
the cat’s tibialis muscle were used (fig. 11 and 12). Certain differ-
ences have been noted, however. Whereas the second of two tetani
in quick succession does not produce a secondary rise in the tibialis

Fig. 11. Pigeon, decerebrated, tibialis anticus. Myogram of a sec. 6 indirect maximal
tetanus at 45 per sec. Time 0.1 sec.

Fig. 12. Pigeon, decerebrated, tibialis anticus. Two 10 sec. maximal indirect tetani
20 sec/ apartat 45 per sec. Single shocks 3 sec. apart. 8 min rest before first tetanus.

TEKSrON CHANGES DURING TETANUS.

215

Pig. 13. Pigeon, decerebrated, tibialis anticus. Upper curve myogram, lovrer curve
action potentials. Tetanus 10" at 40 per sec. Time 0.1 sec. The first, middle and
last part of the curve is shown.

of the cat and is also less effective ivith regard to the production
of a post-tetanic potentiation, the second tetanus in the pigeon
showed a definite secondary rise (fig. 12) under the same conditions.
However, the second tetanus was frequently followed by a slight
depression of the subsequent single twitches.

The action potentials during tetanus sometimes showed a strik-
ing decrease in spike height during the rise in tension (fig, 13),
There was also a definite reduction in the size of the a. p. of the
single twitch following the tetanus in spite of an increase in ten-
sion. Both of these observations are in full conformity with the
results on mammalian muscle. During a second tetanus the de-
crease in voltage of the action potentials occurred more rapidly
than during the first tetanus.

Brown and Harvey (1938) have reported that the second of
two muscular action potentials separated by a short interval in
the gastrocnemius muscle of the hen anesthetized with pemocton,
was larger than the first. This was observed occasionally, but as a
rule there was no difference in the spike height between the first
and the second action potential in a tetanus yielding discrete
action potentials (40 — 45 p, sec.), (fig, 13).

Discussion.

The secondary rise phenomenon described in this paper obviously
bears definite relations to the potentiating effect of a tetanus on
subsequent single muscle twitches as reported by Gutxman ei al.
(1937), Rosenbeueth and Morison (1937), Feng et al. (1938),

216

TJ. S. TON EULER AND ROY L. SWANK.

Brown and Euler (1938) and Pirquet (1938). These relations
may be summarised in the following points.

(1) The secondary rise in tension in a tetanus, elicited by a
suitable indirect or direct stimulation of the tibialis of the cat or
the pigeon is regularly followed by a potentiation of subsequent
single twitches.

(2) In cases where the post-tetanic potentiation is absent no sec-
ondary rise occurs (iniection of KCl in sufficiently large doses; in
a second tetanus within a short time after the foregoing; and in
soleus at low tetanizing frequencies).

(3) The action potential of a single twitch following a tetanus
is usually depressed though the twitch tension may be increased
up to 100 p.c., and the action potential during a secondary rise
show a gradual decrease in spike height. We have observed a
decrease of more than 50 % of the spike height of the action poten-
tial at a time when tension of the muscle was distinctly raised.
It will appear that the altered state in the muscle, which produces
a post-tetanic potentiation functions during tetanus to produce an
increase in muscular tension.

In addition to other factors which influenced the development
of the secondary rise, i. e. frequency of stimulation, and interval
between tetani, the condition of the preparation seems to be
important. We have not been able to find a definite relationship
between the initial tetanic tension and the secondary rise, although,
admittedly, the phenomenon was usually more striking in cases
where rhe initial tetanic tension was not too high. This was not
due to a submaximal stimuli, for an increase in stimulus strength
did not change the configuration of the curve. The initial tetanic
tension at moderate frequencies generally decreases somewhat in
our experiments an hour or so after the preparation of the muscle
and it was then that the secondary rise phenomenon was most
easily demonstrated. This “aging” process also occurred during a
period of rest, with the muscle unhooked and kept warm, and under
these circumstances the effect was often very conspicuous. We feel
that this change cannot be explained solely on a circulatory basis.

There are obviously two possible explanations for the slowly
developing rise in tension during a tetanus, namely, (1) a greater
number of fibres entering into action simultaneously and (2) an
increase in the tension of each fibre. Of course, these events may
appear together. In a few experiments a slight initial increase

TENSION CHANGES DURING TETANUS.

217

occurred in the size of the action potential during tetanus indi-
cating, perhaps, that more fibers were becoming active. The ac-
tion of the tetanus here is clearly the same as is seen in a partially
curarized muscle. On the other hand in most experiments the
action potentials decreased from the beginning of the tetanus.
It would appear difficult to decide if the size of the action poten-
tials gives any definite information as to the number of fibers
partaldng in the contraction.

A decrease in the amplitude of the action potentials during
a tetanus could be due, at least partly, to temporal dispersion. We
have observed a marked reduction in spike height, however, vdth-
out any noticeable change in the width of the action potential,
and have also observed a definite broadening of the potential
with only a small reduction in size. It should also be recalled that
there is often no sign of a an increased temporal dispersion in the
smaller action potential accompanying the potentiated tnitch
which follows a tetanus.

From the beginning of a tetanus the time necessary for a sec-
ondary rise in tension to appear is usually about 2 seconds.
During the first second, there is frequently a slight fall in tension,
without significant changes in the action potentials. The reason
for this dip is not clear, but we may regard it as the interference
minimum of two curves, one falling (fatigue) and one rising (spe-
cific efiect). At any rate we have no evidence for a temporary
reduction in the number of active fibers, since the action potential
is, as a rule, unaltered during this stage. If we compare the first
three contractions of the first, second, and third tetani in fig. 9,
we find that this combined tension is lowest in the second tetanus.
In the third tetanus (not shown in the figure) the first contraction
is already increased, indicating a change in the state of the
muscle though the total tension is still slightly less than was
attained by the first three contractions of the first tetanus. Sub-
sequent to the third tetanus the first contraction in each tetanus
remains slightly elevated, although there is a slight final regression,
probably due to fatigue. It therefore appears that after the two
first tetani a change has taken place in each muscle fiber so that
it contracts with greater force. The total tetanic contraction is
also dependent upon other factors such as the amount of available
contraction material, etc.

In the muscle of the pigeon the period of depression or fatigue
is less obvious and the muscle is able to produce a secondary

218

U. S. VON EOLBR AND ROV L. SWANK.

rise after a very short interval. No doubt this indicates a quicker
restoration to the pretetanic conditions, and it correlates with the
fact that the post-tetanic potentiation lasts for a much shorter
time in the pigeon than in the cat.

In the soleus we have not noted any changes in the action
potentials during the rise in tetanic tension, using low stimulating
frequencies and it thus seems that the mechanism for the rise is
different in this muscle. We have confirmed the observation of
Brown and Euler that the post-tetanic potentiation was less
easily elicited in this muscle. Thus an interpolated tetanus of 2 — i
seconds duration at a moderate frequency did not, as in the tibialis,
lead to an increased tension of the following twitch but, reversely,
to a decrease. This is in keeping with our observations that a series
of short tetam' of the soleus result in a gradual fall in tension of this
muscle, whereas a complete recovery is obtained in some 5 or 10
seconds when suitable tetanming frequencies are used. When much
higher frequencies were used even a short tetanus on soleus often
produced a post-tetanic potentiation, as shown by Eeng et al.
(1938). On these occasions there was also a clear secondary rise,
which strengthens our view that these two phenomena are causally
related.

On similar grounds as those presented by Brown and Euler for
the post-tetanic potentiation we cannot consider the secondary
rise similar to the Orbeli phenomenon.

The expenses of this investigation have been defrayed by a
grant from the Therese and Johan Andersson Memorial Foundation
to one of us (U. S. v. E.).

Summary.

Indirect maximal tetanic stimulation of the tibialis muscle
of the cat and the pigeon at frequencies of about 45 p. sec. fre-
quently lead to a gradual increase in tension of the muscle to a
considerable extent.

This secondary rise in tension is accompanied by a diminution
of the action potentials. When a series of short tetanic stimulations
are applied to the nerve the maximal tension of each tetanus shows,
after a short depression, a gradual increase up to a maximum fol-
lowed by a fall.

TENSION CHANGES BORING TETANUS.

219

In tlie soleus muscle of the cat this phenomenon is much less
obvious at frequencies of 15 — 40 p.s. and there are only slight
changes in action potentials at these stimulating frequencies.

The phenomenon described here appears to be closely related
to the post-tetanic potentiation of a single muscle twitch.

Eeferences.

Brown, G. L., J. Physiol. 1937, 91 , 4P.

Brown, G. L., J. Physiol. 1938, 92 . 22P.

Brown, G. L., and U. S. v. Euler, J. Physiol. 1938, 93 . 39.
Brown, G. L., and A. j\L Harvey, J. Physiol. 1938, 93 . 285.
Feng, T. P., L. Lee, C. Meng and S. Wang, Chin. J. Physiol. 1938.
18 . 79.

Guttman, S. a., R. G. Horton and D. T. Wilbur, Amer. J. Physiol.
1937, 119 . 463.

hlATTHEWS, B. H. C., J. Physiol. 1938. 93 . 25P.

PiRQUET, A. V., Pfliig. Arch. ges. Physiol. 1938, 240 . 763.
Bosenblueth, a., and R. S. Morison, Amer. J. Physiol. 1937. 119 .
236.

Weber, A., Die Electrocardiographic, Springer, Berlin 1939.

16 — i01323. Acta phys. Scandinav. Vol. 1.

Aus dem Physiologischen und Chemischen Institut
der Konigl. Veterinarhochschule zu Stockholm.

Pyi’opliosphatase im BlutA

Von

KNUT SJOBERG.

Das Enzym, Pyrophosphatase, welches anorganische Pyro-
phosphate und organisch gebundenes Pyrophosphat in anorga-
nische Orthophosphate uberfiihrt, wurde von EAy in verschie-
denen Organen und in kleiuer Menge im Blut nachgewiesen. Auch
Roche konstatierte das Vorkommen eines derartigen Enzyms
in den Blutkorpern und im Serum, Irgendwelche Angaben fiber
die Variationen der Enzymaktivitat bei verschiedenen Tierarten
und unter verschiedenen Umstanden bei Tieren wurden jedoch
nicht gemacht.

Das Vorkommen verschiedener Stoffe im Blut kann tells dar-
auf beruhen, dass diese daselbst eine gewisse Funktion zu erffil-
len haben, teils dass sie intermediare Stoffwechselprodukte bil-
den, die mit Hilfe des Blutes zwischen den verschiedenen Gewe-
ben und • Organen transportiert werden. Letzteres gilt besonders
fur gewisse im Serum vorkommende Yerbindungen. Da sich
die Pyrophosphatase, wie spater gezeigt werden wird, haupt-
sachlich in den Erythrozyten vorfindet, scheint dieselbe bei den
Stoffwechselprozessen in diesen eine gewisse Aufgabe zu haben.
Lohmahn hat nachgewiesen, dass das Blut und besonders die
Erythrozyten organische Pyxophosphatester in Form von Ade-
nylpyrophosphat enthalten, das als Substrat ffir die Pyrophos-
phatase dienen kann. Sicher kommen auch andere organische
Pyrophosphorsaureester vor. So wird ja beispielsweise das Aneurin
an Pyrophosphat gekoppelt xmd bildet Cocarboxylase.

Mit Rucksicht auf die grosse Bedeutung, die derartigen Pyro-
phosphatestern und deren Synthese und Hydrolyse zukommt.

* Der Redaktion am 6. August 1940 zugegangen.

rYROPUOSPHATASE IM BLUT.

221

wird liier iibcr eine Metliode zur Bestiramung von Pjnrophos-
phatasc im Blut und Blutserum bez\r. Plasma nebst einer vor-
liinfigon Untcrsucbung iiber das Vorkoramen dieses Enzyms im
Blut dcs jMensclien und verschiedener Tierarten bericlitet. Im
Zusammenbang hiermit Avurde die Abbangigkeit der Pyropbos-
pbatase von der Wasserstoffionenkonzentration und ihre Veran-
derungen bei der Aufbevrahrung der Blutproben in vitro unter-
sucbt. Schliesslicb vurdcn die Pyrophospbat- und Pyropbos-
pbatasemengcn im Blut miteinander verglicben.

Yersnclismctliod I k.

Als Substrat wurdc Xatriump)Topbospbat angewandt, und die
Enzymvdrktmg vurdc durcb Bestimmung des wabrend einer
gewissen Zeit freigewordenen Ortbopbospbats nacb Fiske und
SuBBAROW festgestellt. Da grossere Mengen IsatriumpjTropbos-
pbat die Bildung der blauen Vcrbindung bemmten, die bierbei
alls jVmrooniumpbospbormolybdat und dem Keduktionsmittel
cntstebt, musstc die !Menge des Pyrophospbats in der Beaktions-
misebung relativ niedrig gcbaltcn ivcrden.

Folgcndc Reaktionsraisebung erodes sicb als geeignet:

Na4PsO.-L6sung 4 ml

MgSO^-Losung 4 ml

Boratpuffer 10 ml

Blut bezv. Serum oder Plasma ... 4 ml

Aq. dest 18 ml.

Zusammensetzung der Losungen:

NatPiOrLosung. 4.000 g NaiPaO, + 10 Aq. fiir analytisebe
Zweeke werden in 900 ml dest. Wasser gelost. Zusatz von so
viel HCl, dass pH = 7.4 vdrd, danacb Verdiinnung auf 1 000 ml
mit dest. Wasser. Die Losung zerteilt sicb allmablicb bei der
Aufbewabrung. tlberscbicbtet man die Losung mit Petroleum-
atber und bewabrt sie bei niedriger Temperatur auf, so kann
sie sicb einige Woeben lang brauebbar balten.

MgSOtrLosung. 5.00 g reines MgS04 + 7 Aq. werden in 1 Lit.
dest. Wasser gelost.

Boratpuffer nacb Palitzsch pH = 7.4. 10 Teile M/20 Borax +
4- 90 Teile M/6 Borsaure, M/20 NaCl.

Die Bestimmung wd bei 37—38° C. im Wassertbermostaten
ausgefiibrt. Nacb gewissen Zeitintervallen, zweekmassig nacb

i,^nz ym vtrkun^

222 KNDX SJOBERG.

30, 60, 120 Min., werden Proben von 10 ml entnommen nnd in
10 ml lOptoz. Tricliloressigsanre pipettiert, danach, wird fil-
triert. In 10 ml des Piltrats wird die Menge anorganiscben Ortho-
phosphat-P nach. Piske nnd Subbasow mit Hilfe eines photo-
elektrischen Kolorimeters bestimmt.

In einer besonderen Probe stellt man die urspriingliche Menge
anorganiscben Orfcbopliospbat 7 P im Bint fest.

Als Mass fiir die Enzymmenge berecbnet man die Reaktions-
konstante nacb der Formel fiir monomolekulare Reaktion

wo t die Anzabl der Minuten zwischen dem Beginn der Reaktion
und der Zeit fiir die Probenentnahme bedeutet, a die totale'ur-
spriinglicbe Menge Pyropbosphat-P in der Reaktionsmischung und
X die wabrend der Zeit t gespaltene Menge Pyropbospbat-P. Bei
der Entnabme einer andern Blutmenge als der pben vorgescblage-
nen, was besonders bei starker Enzymwirksamkeit zweckmassig
sein kann, recbnet man zu 1 ml Blut bezw. Plasma pro 10 ml
Reaktionsmiscbung um.

Man kann aucb in Analogic mit der von Bodansky ange-
gebenen Einbeit fur Ortbomonopbospboresterase diejenige Menge

PrKOPHOSPHATASE IM BLET. 223

au auorganischem Orthophosphat-P feststellen, welclie unter den
oben angegebenen Bedingungen vrahrend 1 Stunde aus 100 ml
Blut bezw. Plasma gebildet wird. Dieser Wert, der bier mit
Pyro-E bezeichnet wird, stellt sicli natiirlicb weniger exakt als
die Keaktionskonstante, da die Spaltung nur eine kurze Zeit
geradlinig verlauft, er diirfte jedoch fur klinische Bestimmungen
ausreicben.

Die Bestimmungen wurden mit Blut ausgefiihrt, dem Natrium-
zitrat zugesetzt -war.

pH-Optinium.

Kay gibt als pH-Optimum fur die Blutpyropbospbatase 7.6
und Roche 6.2 an. Da diese Werte ganz bedeutend divergieren,
liielt ich es fur notig, dieselben zu kontrollieren. Das Resultat
findet sicb in Fig. 1. Als Optimum erbielt ich pH = 7.4, also
einen Wert, der mit dem von Kay angegebenen fast iiberein-
stimmt.

Der Pyropliospliatasegeliftlt im Blut.

Tabelle 1 gibt eine Zusammenfassung der erhaltenen Werte
fur die Pyrophosphatasevrirkung, sowohl in Form von Reaktions-
konstanten als auch von Pjto-E (in der Tabelle nur mit E be-
zeichnet) angegeben. Im Plasma ist die Pyrophosphatasewir-
kung relativ gering und zeigt bei verschiedenen Tierarten keinen
statistisch sicheren Unterschied. Der Pyrophosphatasegehalt in
den Erj’tlirozyten wurde aus den Werten fiir Blut und Plasma
mit Hilfe der Hamatokritwerte berechnet. ]\Iit Riicksicht auf
die Pyrophosphatasewirkung der Erythrozyten kann man zwei
Gruppen unterscheiden. Die eine umfasst Blutkorper von Mensch,
Hund und Pferd mit relativ kraftiger Wirkung, die andere Blut-
korper vom Rind mit einer Enzymwirkung, die nur ungefahr
^/jo der Wirkung in den erstgenannten betriigt. Es kommen je-
doch sehr grosse individuelle Variationen, besonders im Rinder-
blut, vor. Die Werte fiir die PjTophosphatasewirkung im ganzen
Blut folgen nicht den Werten in den Erythrozyten, da das Blut-
korpervolumen bei verschiedenen Tierarten variiert. Dasselbe
ist beim Pferde mit durchschnittlich 30 am niedrigsten, beim
Menschen, Hunde und Einde findet man normale Werte zwi-
schen 40 und 50.

In die Tabelle wurde auch das Verhaltnis z-wischen den Werten
aufgenommen, die man nach den beiden Berechnungsmethoden

224

KNOT SJOBBRG.

Erythrozyten

100 k E 100 k E 100 k E ,

XUU k jQQ jj. lUO k

109 118
103 129

128 135
165 129
66 120
59 111
86 125
122 138
133 no

Mittel
mittl. Feil.
Eel.

Mittel

mittl. Fehl +
Eel.

Mittel

mittl. FehL ±
Eel.

6.76

161

10.9

168

8.10

133

8.35

137

9.26

165

18.1

150

8.90

160

PYROPHOSPHATASE IM BRUT.

225

erlialt. Will man nur die AVerte von ein und derselben Tierart
vergleiclien, so liefern die Pyxo-E-Ziffern \vemgstens fiir die Hi-
niscbe Beurteilung brauclibare AVerte. Dagegen Yrird der Unter-
scMed grosser, ■wenn es sicb. nm. einen A^ergleicb zvdscben der
PyrophospbataseTvirlaing verschiedener Grossenordnung han-
delt.

Die fiir die Pyrophosphatasewirkung in Blut und Erytlixo-
zyten erbaltenen A\’’erte entsprechen jedoch nicht der tatsacb-
lichen Enzynavirkung, da das Plasma — wie spater gezeigt
vrerden wird — eine hemmende Substanz entbalt. Die angege-
benen AA^erte driicken also den Einfiuss aus, der in bamolysiertem
Blut zur Geltung kommt. AA^unscbt man einen Masstab fiir die
tatsacbbcbe Pyropbospbatasewirkung in den Erytbrozjdien, so
muss die Bestimmung direkt an diesen nacb der Beseitigung des
Plasmas ausgefiibrt werden. Uber diese Bestimmungen wird
spater in dieser Mitteilung bericbtet werden.

Der Pyropbosphatgebnlt im Blut.

Der Pyrophospbatgebalt wird nacb Lohmann durcb Erbitzung
von Blut bezvr. Plasma mit dem gleicben Volumen 2 n HCl 7
Minuten lang bei 100° C. bestimmt, wonacb man den freigewor-
denen anorganiscben Pbospbor auf gewbbnbcbe Weise fest-
stellt.

Tnbelle 2.

PyrophospJiaie in mg pr 100 ml.

Anzahl

Probe

Plasma

Erythrozyten

Variation

Mittel

Variation

Mittel

Homo . . .

0,10—1.17

0,70 ± 0.18

7.26—12.60

9.81 ± 0.97

Hund . . .

O.OO— 1.71

0.78 ± 0.86

3.84— 7.66

5.64 ± 1.10

Pferd . . .

0.1 0— 0.66

0.31 ± 0.06

1.90— 5.86

4.00 ± 0.40

Kind . . .

O.OO— 0.71

0.24 + 0.06

2.42— 6.20

3.74 ± 0.22

Tabelle 2 gibt die von mir erbaltenen AVerte fur den Pyropbos-
pbatgebalt im Plasma und in den Erythrozyten von verscbiede-
nen Tierarten an. Solcbe Bestimmungen wurden vorber bei-
spielsweise von Kerb und Daotjd ausgefiibrt, deren AVerte im
■grossen ganzen mit den meinigen ubereinstimmen. Offensicbt-
lich entbalt das Plasma relativ geringe Mengen an Pyropbos-

226

KNUT SJObERS.

phaten, und ein deutUcher Unterscliied bei verscbiedenen Tier-
artea vrurde nicbt erzielt. Der durciiweg niedrige Pyiophos-
phatasegehalt im Plasma diirfte mit der geringen Menge der da-
selbst vorhandenen Pyxopbosphate im Zusammenhang steben.
Die Exytbrozyten ia Menscbenblut entbalten ungefabi doppelt
so viel Pyxopbospbate 'wie die Blutkorpex dex iibxigen Tiexaxten.
Biytbxozyten vom Menscben eigaben aucb bobeie Pyxopbos-
pbatasewirkung. Es liegt jedocb keine direkte Konelation zwi-
scben Pyxopbospbatgebalt und Pyxopbospbatasewirkung vox.
Letzteie mrd offenbai duicb andeie Eaktoxen beeinflusst.

In diesem Zusammenbang kann beivoxgeboben "wexden, dass
dex Gebalt an andeien oxganiscben Pbospboxsaureestein in Blut-
koxpein von Menscben, Hund und Pfexde im Duxcbscbnitt im-
gefabi lOmal bobei ist als in Blutkoxpern vom Rinde.

Das Verhalten der Pyrophosphatase im Blut in vitro.

Die Bestimmung dex Pyxopbospbatasevrirkung gemass dex
oben bescbxiebenen Metbode wixd in einex Reaktionsmiscbung
ausgefubrt, die im Veibaltnis zum Blut bypotoniscb ist, vresbalb
die Exytbxozyten bamolysieit wexden. Da die Pyxopbospbatase
in den Blutkoxpexn voxkommt, kann die FeststeUung von In-
teresse sein, ob dieselbe ihie "Wirkung auf die Reaktionsmiscbung
auszuiiben vexmag, obne dass eine Hamolyse stattfand. Dieses
ist nambcb mit dex Pbospboxmonoesteiase der Fall, wie icb in
einex friibexen blittedung gezeigt babe.

Es wurden desbalb einige Versucbe mit Pferdeblut ausgefubrt,
in dem die Pyxopbospbatasewirkung auf die vorbex bescbriebene
Weise, docb mit dem Unterscbiede bestimmt wxirde, dass die
Reaktionsmiscbung ausser ibren gewobrdicben Bestandteilen
aucb 0.8 Prozent HaCl entbielt. In dieser bliscbung findet keine
Hamolyse statt. Der Versucb wnirde teils mit Blut und teils mit
Blutkorpexaufscblammung ausgefubrt.

Letztere vnixde auf die Weise bexgestellt, dass nacb dex Zentri-
fugierung von Zitxatblut das Plasma abpipettiert, die Blutkorpex
zweimal mit pbysiologiscber Kocbsalzlosung gewascben und
scbliessbcb in so viel pbysiologiscbex Kocbsalzlosrmg aufge-
scblammt WTixden, dass das Volumen ungefabr die Halfte des
uxspriinglicben Blutvolumens betrug.

Da die Bestimmung der Pyxopbospbatasewirkung in isotom-
scber Reaktionsmiscbung vorgenommen wnirde, trat nur eine

PYROPHOSPHATASE IM BLOT. 227

TaTjcllo 3.

Pyrophosphatasewirkung in Pyro-E tin Pfcrdehlut.

Blnt-

probe

Nr.

Tage nach
Blntcnt-
nahme

Normale

Reaktions-

mischnng

Reaktioiis-
mischnng +
0.8 % NaCl

Bejnerknngen

8.B

Bint

7.5

Bint

iy.8

5.5

Bint

19,8

8.4

19.6

11.2

Hamolvse

25.8

21.0

nabc vollstknd. Samolyse

Bint

etwas H&molvse

11 3

kraftige Hamolyse

Plasma

Erythrozytcn

Bint

etwas Hamolyse

Hamolyse

5.3

Plasma

6.9

6.4

8.4

114

Erytbrozyten

103

125

34.8

Bint

35.8

27.9

Hamolyse

6.4

Plasma

5.1

7.6

Erytbrozyten

123

175

BlntkSrp.-anfschlammung

175

Hamolysierte >

Blntk6rp.-anfschl8mmnng

142

Hamolysierte »

228

KNUT SJOBERQ.

Blnt-

probe

Nr.

Tage nacb
Blntent-
uahme

Nonnale

Beaktions-

miscbung'

Eeaktions-
mischnng +
0.8 % NaCl

Bemerkungen

342

Blntkorp.-aufschlanimting

422

4.6

414

14.1

etwas Hamolyse

406

10.3

Hamolyse

D:o + Plasma

217

Blntborp.-aufschiammnng

261

393

Hamolyse

D:o + Plasma

270

Blntkorp.-anfschlammung

388

992

276

Hamolyse

unbedeutende Spaltung des Natriumpyropbospbats ein, welcbe
Spaltung dem Gebalt des Plasmas an Pyropbospbatase zuzu-
scbreiben ist. In der Blutkorperaufscblammung Hess sicb unter
diesen Versucbsbedingungen keine Enzymwkung nacbweisen
(Tab. 3). Die Pyropbospbatase befindet sicb also in den Erytb-
rozyten in einem solcben Zustande, dass sie mit einem Substrat
ausserbalb diesen nicbt in Kontakt zu kommen vermag. In dieser
Beziebung begt zwiscben diesem Enzym und der Pbospbormono-
esterase ein wesentbcber Unterscbied vor.

Wird die oben genannte Blutkorperaufscblammung bamoly-
siert imd fdtriert man die Stromata ab, so zeigt das Hamolysat
die gleicbe Aktivitat wie die entsprecbende Menge Blutkbrper-
aufscbwemmung in bypotoniscber Reaktionsmiscbung. Die Py-
ropbospbatase ist folglicb nicbt an die Stromata gebimden (Tab.
3, Nr. 3 b und 4 b).

Bei der Bestimmung der Pyropbospbatasewirkung in der
Blutkorperaufscblammimg in bypotoniscber Losung wurden be-
deutend bobere Werte als diejenigen erzielt, welcbe aus den Wer-
ten fur Blut imd Plasma berecbnet wurden (Tab. 3). Dieses
deutet darauf bin, dass das Plasma irgendeiue Substanz ent-
balt, die die Pyropbospbatasewirkung bemmt.

Zwecks Untersucbimg dieses Verbaltens wurde folgender Ver-
sucb gemacbt. Die Pyropbospbatasewirkung einer Blutkorper-
aufschlammung wrirde auf die gewobnlicbe Art bestimmt.
Gleicbzeitig wurde eine andere Probe mit derselben Blutkorper-
aufscblammung (2 ml) nebst Plasma (4 ml) angesetzt. Im letz-

PYROrapSPHATASE IM BLUT.

229

teren Falle erliielt ich einen bedeutend niedrigeren Wert fiir die
Tyrophospbatasewirkung (Tab. 3, bfr, 5 b und 6 b). Hieratis
geht also dcufh'ch hervor, doss das Plasma ehie hemmende Substanz
enihdlt.

Ein vreitercr, bier bebandelter XjBistand ist die Frage, ^vie sicb
die Pyropliospbatasew-irkung verandert, -vrenn eine Blutprobe
einige Tage in vitro anfbewabrt wird, bis eine Hamoljse in der-
selben cinsetzt.

Z^vecks XJntersucbnng dieses Yerbaltens wurde teils Pferde-
blut, teils Blutkorperaufschlannnung bei Zinimertemperatur
(20 — 23° C.) 1 — 3 Tage lang aufbewabrt und biernacb die Pbos-
phatasewirkung sowohl in gewobnlicber Reaktionsniischung als
aucb in solcber init Zusatz von 0.8 Proz. NaCl bestimnit.

Wurde die Bestininmng init Blut in gewobnlicber Losung aus-
gefubrtj so erbiclt icb nacb 2 — 3 Tagen, wabrend welcber Zeit
eine starke Hiimolyse eingetreten war, keine sicbere Anderung
in der Pyxopbospbatasewirkung. Dagegen nabm dieselbe etwas
im Plasma zu. Bei der Bestimmung in isotoniscber Losung stieg
die Pyropbospbatasewirkung entsprecbend dem Grade, in wel-
chcm die Hiimolysc fortschritt und ein Teil des Enzyms ins
Plasma tiberging.

In Blutkorperaufscblamniungen nabm die Pyropbospbatase-
wirkung schon nacb 24 Stunden zu, wenn die Bestimmung in ge-
wohnlicber Beaktionsniiscbung ausgefulurt wurde (Tab, 3). Bei
langerer Aufbewabrung (2 — 3 Tage) wurde in gervissen Fallen
Avieder eine Yerminderung der Pyropbospbatasewirkung gefimden,
was auf einer Inakti^’ierung des Enzyms beruben durfte. Beim
Yersucb in isotoniscber Losung wurde — wie oben genannt — in
einer friscben Probe keine Enzymtatigkeit gefunden, nacb 1 — 2
Tage langer Aufbewabrung konnte aber in einem Falle eine
scbwacbe IVirkung hacbge^viesen werden. Die Ursacbe bierfiir
muss in einer beginnenden Hamolyse gelegen baben.

Die Pyropliosphatasewirknng in Toni Plasma befreiten

ErytUrozyten.

Die Pyropbospbatasewirkung wurde in Blutkorperaufscblam-
mungen bestimmt, die auf die oben angegebene Weise bereitet
waren. Die Feststellung des Blutkbrpergehalts gescbab nacb der
Hamatokritmetbode .

230

KNUT SJOBERG.

Tabelle 4.

Pyrophosphatase in den Erythrozyten.

100 k, aus den IV’erten

100 k, direkt

n/r

fur Blut u. Plasma ber.

bestimmt

1 Homo

1.28

6.95

5.4

0.88

8.65

9.8

0.95

9.20

9.7 ,

Rel. Hit. 15

Hittel 8.3

Hund

0.9 0

2.13

2.4

0.65

2.54

3.9

Efel. Hit. 11

Hittel 3.2

Pferd

0.48

8.65

. 0.91

10.50

0.70

12.20

Rel. Hit. 10

Hittel 16

Rind

0.061

0.112

1.8

0.042

0.124

3.0

0.061

0.132

2.2

0.121

0.178

1.5

Rel. Hit. 1

Hittel 2.1

In Tabelle 4 finden sich die fiir die Pyrophospliatasewirkung
in den Erythrozyten erhaltenen Werte, die teils aus den Werten
fiir Blut und Plasma berechnet, teils direkt in der Blutkorper-
aufschlammung bestimmt wurden. Die letzteren AVerte mirden
dahin umgerecbnet, dass sie fiir das gleiche Blutkorpervolumen
■wie die erstgenannten gelten, also 4 ml auf 40 . ml Reaktions-
mischung.

In samtbchen Blutarten erhielt ich nach der letztgenannten
Methode bedeutend hohere AVerte. Bei Hinder-, Hunde und
Menschenblutkorpern belief sich die Steigerung abgerundet auf
das 2-, 3- bezw. 8-Fache, bei Pferdeerythrozyten war die AVir-
kung 16mal starker. Hieraus geht hervor, dass das Pferdeblut
besonders reich an dem hemmenden Stoff ist.

Der Gehalt der Erythroz 3 rten an Pyrophosphatase bei dem Men-
schen und verschiedenen Tierarten variiert ganz bedeutend und
gibt fiir die untersuchten EaUe steigende AA^erte in der Eeihen-
folge: Rind, Hund, Mensch, Pferd.

PYROPHOSPHATASE IM BLUT.

231

Ziisammenfassung.

1. Eine jMethode zur Bestimmung der Pyrophosphatase-
wirkimg in Blut bezw. Plasma, Serum und Erytlirozyten wird
beschriebcn.

2. Das pH-Optimum fur die Pyrophospbatase\Yirkung im
Blut liegt bei 7.4.

3. Die Pyropbospbatasewirkung wurde im Blut und Plasma
bei jMenscb, Hund, Pferd und Rind bestimmt. Mit Hilfe dieser
Werte und der Hsimatokritwerte wurdc die Wirkung in den Eryth-
rozyten berecbnct. Blut der beiden erstgenannten entbalt un-
gefabr lOmal mebr Pyropbospbatase als Rinderblut, und Pferde-
blut ungefabr 5mal so viel. Die Pyropbospbatase kommt baupt-
Sficblicb in den Erjdbrozyten vor, Plasma entbalt nur unbedeu-
tende i^Iengen.

4. AVird die P 3 ’Topbospbatasewirkung direkt in vom Plasma
befreiten Er3’ibrozytcn bestimmt, so crbiilt man bedeutend hohe-
re Werte als bei der Berecbnung der Wirkung aus den Werten
fiir Blut und Plasma. Die Zunabme variiert zwiscben dem Dop-
peltcn und 18-Eacbcn. Die Enz 3 Tnwirkung in den Er 3 ’^throzyten
verhiilt sich bei Pferd, bicnscb, Hund und Rind wie 76 : 57 :
17; 1.

5. Irgendeine Korrelation zwischen dem Gebalt der Blut-
korper an P 3 rrophospbaten und P 3 ’Topbospbatase liegt nicht vor.

6. Wird die Bestimmung der P3T:opho5phatasevrirkung in
einer Reaktionsmiscbung ausgefiibrt, die mit den Blutkorpern
isotoniscb ist, so erbalt man keine Enz 3 rmwirkung. Die P 5 t;o-
pbospbatase in den Blutkorpern kann folglicb nicbt ausserbalb
diesen zur Wirkung gelangen, bevor die Erythrozyten bamoly-
siert sind. Hierin unterscbeidet sich die P 3 rrophospbatase von
der Pbosphormonoesterase.

7. Plasma bezw. Serum entbalt einen Stoff, der die Pyro-
phospbatase'wirkung stark bemint. Dieser Umstand erklart das
in Punkt 4 genannte Verbalten.

8. Bei der Aufbewabrung von Blutproben bei ungefabr 20 0.
fiir eine Zeit bis zu 3 Tagen iindert sicb die Pyrropbospbatase-
vrirkung nur unbedeutend. In vom Plasma befreiten Blutkorper-
aufscblammungen nimmt die Pyuropbospbatasewirkung nacb 24
Stunden langer Aufbevrabrung bei 20° G. zu.

232

KNUT SJOBERG.

Literatur.

Fiske, C. H. and Y. Sotbarow: J. biol. chem. 1925, 66 , 375.
Kay, H. D. Biochem. J. 1928, 22 , 1446.

Kerr, S. E. and L. Daoud. J. biol. chem. 1935, 109 , 301.
Lohmann, K. Biochem. Z. 1928, 202 , 466.

— , Ebenda 1928, 203 , 164.

— , Katurwissenschaften 1929, 17 , 624.

Boche, j. Biochem. J. 1931, 25 , 1724.

Sjoberg, K. Diese Z. 1940 (In print).

From the Biochemical Department, Karolinska Inatitutet, Stockholm.

Traiisainination Peptide Substrates
in Cattle Diaphragm Muscle.^

GUNNAR AGREN.5

The disappearance of added glutamic in pigeon breast muscle
was first discovered by Neebiiam (1930), while the concentration
of van Slyke nitrogen remained stable. From the results obtained
she concluded that the amino group of glutamic acid is trans-
ferred to a reactive carbohydrate compound in order to form a
new amino acid. Braunstein and Kritzmann (1937, 1938, 1939)
found the explanation in the reaction:

Glutamic acid -f* a-ketonic acid o-ketoglutaric acid -}- amino

acid

which was isolated and examined.

The reaction is reversible and speedy under certain conditions.
It has been found in most animal tissues and the reaction is
termed "transamination” (Braunstein and Kritzmann, 1937).
The frequency of transamination and the rapid rate at which it
occurs suggest the reaction playing a prominent part in inter-
mediary tissue metabolism, the significance of which being as
yet not determined. In the present paper the problem is attacked
from a different angle: The possible formation of dipeptides in
a reaction between glutamic acid and a-ketonic acids in peptide
linkage with an amino acid. A reaction including the disappear-
ance of Van Slyke nitrogen in liver extracts has been scrutinized
in a series of experiments (Agren c. s. 1937, 1939; Agren, 1940 a).

* Received for publication 29 June 1940.

~ Fellow of the Rockefeller Foundation 1938 — 1939.

234

GUNNAR AGREN

One of the initial steps in the process was apparently a reaction
between amino and aldehyde groups (amino acids and glucose),
in which compounds were formed similar in type to the Shiff
bases. The theory was advanced that, in a reaction between
ketoaldehydes and amino acids, a compound could be obtained
which would develop into a peptide after dehydrogenation and
transamination. The rapid interaction between methylglyoxale
and amino acids was demonstrated in a recent paper (Ageen
1940 b). In the present paper one of the conclusive steps is
described by help of experiments, i. e. the transamination of
ketonic acid-amino acid compounds.

Procedure and Methods.

General procedure. In studying the transamination it is ex-
pedient that this reaction is separated from others comprising
amino and a-ketonic acids. Many of these lateral reactions re-
quire molecular oxygen and could, therefore, eventually be
eliminated by working under anaerobic conditions. According to
rule the results obtained by this method should not differ from
those acquired aerobically in the presence of bromoacetate and
sodium arsenite (Braunstbin and Kritzmann, 1937). These
substances are inhibitors preventing certain lateral reactions.
Neither of these methods could completely eliminate the lateral
reactions in cattle muscle, as glutamic acid disappeared slowly,
anaerobically as well as aerobically, in the presence of bromo-
acetate and sodium arsenite, without any addition of ketonic
acids. This lateral reaction may be due to transamination of
free or bound ketonic acids provided by the muscle. The rate of
reaction is, however, so slow that it does not interfere with the
transamination of added ketonic acids.

A study of transamination was made by the determination of
glutamic acid, formed either by adding a-ketoglutaric acid, amino
acids and dipeptides, or made to disappear by adding glutamic
acid and free or bound a-ketonic acids to muscle suspensions. In
each experiment controls were made with a-ketoglutaric acid or
glutamic acid alone in order to distinguish the unspecific forma-
tion of glutamic acid from preformed amino group donators
(NHs and amino acids) or the consumption of glutamic acid by
preformed ketonic acids. The procedure taken by Braunstein
and Kritzmann (1937) in the determination of glutamic acid is

TKANSAMIXATION Ih’ CATTLE DIAPHIUGM MUSCLE. 285

not Specific and other controls were necessary. In many of the
transamination experiments the glutamic acid precipitate is con-
taminated with some of the a7nino group donating substrates
(amino acids, peptides). In all e.xperiments, therefore, blank
determinations were carried out by adding amino group donators
and receptors to the heat inactivated muscle suspension. Tests
were also made without adding donator and receptor substances.

A considerable variation wa.s apparent in the activity of the
transamination sj-stein in muscle s\ispensions from different
animals. Consequently the activity of the transamination enzymes
in the experiments with peptides and a-kctonic acid-amino acid
compounds were controlled by following the reaction between
alanine and a-kctoghitaric acid or pjTiivic acid and glutamic acid.

Experimental procedure. Cattle diaphragm muscle was chilled in
the slanghter-housc immediately after the death of the animal,
and finely cut in a mincer. The minced muscle was suspended in
4 parts of 0.2 per cent KHCO3, the substrates and inhibitors being
added as neut ralized solutions. Usually 100 — 200 n Mol. of sub-
strate per gram of muscle was added. Substrates such as glycyl-
aminobcnzoic acid, soluble with difficulty, were added in solutions
saturated at 40“ to muscle suspensions of the same temperature.
The experiments were carried out in Thimberg tubes, anaerobic
or aerobic conditions being maintained by filling the vessels with
Nj or 0;. The tubes were sliakcn at 40° for 30 minutes, unless
otherw'ise stated. In the space of time mentioned tbe transamina-
tion was complete. At the end of the experimental period the
solutions were speedily licated to 100“ and then chilled again.
Trichloracetic acid was added to a final concentration of 5 per
cent.

Analytical procedure. Glutamic acid w'as determined by the '
Jones and Moeller method (1928) in the modification of Beaun-
STEiN and Kritzmann (1937), 3 cc of trichloracetic acid filtrate
being used in each analysis. The values are expressed as mg van
Slyke nitrogen in glutamic acid per cc filtrate. Ammonia was
determined according to Conw'ay and Byrne (1933), total amino
nitrogen by the van Slyke method and a-ketonic acids by titration
with cerium sulphate (Fromageot and Desnuelles, 1933).
Ketonic acid was calculated from the titration values as mg/cc
of G-ketoglutaric acid or pyruvic acid, depending on the substrate
used, a-ketoglutaric acid was synthezised according to the
Neuberg and Ringer method (1915).

17 — i01323. Acta phys. Scandinav. V 0 I.I.

236

QUNNAR AGREN.

Results. The reaction between a-ketoglutaric acid and alanine
and the reversibility of this reaction was first investigated. The
results of one of the series of experiments are given in table 1.
Muscle from one animal was used in each series of experiments.
The transaminating enzyme is called aminopherase in accordance
with Bbaunstein and Kritzmann (1939).

Table 1.

Anaerobic transamination with glutamic acid + pyruvic acid and
cL-lcetoglutaric acid + alanine.

Substrate added

Total amino-N
in mg/cc

Amino-N in
glutamic acid
fraction in
mg/cc

A in
glutamic
acid in
mg/cc

Ketonic acid
in mg/cc

0 min.

30 min.

0 min.

30 min.

0 min.

30 min.

rt-ketoglutaric acid +
alanine

0.71

0.17

0.30

+ 1.87

3.1

3.0

«-ketoglutaric acid . .

0.16

0.71

0.072

0.084

+ O.IS

3.1

Alanine

0.71

0.108

0.097

- 0.11

0.2

Blank

0.16

0.17

0.068

0.070

+ 0.02

0.8

0.2

Glutamic acid + pyru-
vic acid

0.32

0.24

0.19

- 0.63

3.2

3.1

Glutamic acid ....

0.31

0.3J

0.20

0.19

- O.IO

0.3

0.2

3 g. muscle + 1.2 cc of 2 per cent KHCO 3 + water; total volnine 15 cc.
Sodium arsenite m/100, bromoacetate 1 : 5000. Glutamic acid 2 mg/cc. + pyruvic
acid 3 mg/cc. «-betoglutaric acid 3 mg/cc + 1 (+) — alanine 3.6 mg/cc.

The reversibility of the transamination is clearly demonstrated.
Starting with alam'ne and a-ketoglutaric acid, the glutamic acid
formation corresponded to a transamination of 21 per cent of added
NHz-nitrogen. In the opposite direction 22 per cent of the added
glutamic acid was lost. The character of a true equilibrium is not
demonstrable, since the plateaux reached from both sides of the
reaction lie at different levels. It may here be noted that the
end point (or difference in glutamic acid) was independent of the
absolute amounts of reacting substrates. Starting with 50/iMol.
glutamic acid or 100 wMol. a-ketoglutaric acid per gram of muscle,
and with a surplus of pyruvic acid or alanine in both experiments,
the result obtained was a 20 per cent reaction. There was no
change in total van Slyke nitrogen, nor any loss of ammonia. In
contrast to the results obtained by Bragnstein and Kritzmann

TRANSAMINATION IN CATTLE DIAPHRAGM MUSCLE. 237

(1937) the concentration of ketonic acids was unchanged. Keto-
glutaric acid seemed less receptive towards lateral reactions in
cattle muscle than in pigeon breast muscle.

Possibilities of reaching an equilibrium in the reaction:

glutamic acid -f pyruvic acid a-ketoglutaric acid + alanine

appeared to vary in muscles of different animals. Braunstein
and Kritzmann (1937) reported the equilibrium constant to be 1.
In the experiments with cattle muscle usually about 30 to 50
per cent of the added glutamic acid disappeared; only 5 experi-
ments of 40 were negative. In the opposite direction (a-keto-
glutaric acid -J- alanine) the results varied considerably. In the
attempts to elucidate this irregularity in the reaction the influence
of freezing the muscle was examined. The muscle was chilled as
usual in the slaughter-house with a salt-ice mixture for the
transport to the laboratory, the whole proceeding taking about
15 minutes. On arrival at the laboratory the muscle usually had
a temperature of -f 2°. After mincing a part of the muscle was
immediately taken to experiments, the rest of the nainced muscle
was frozen at — 20° for 3 hours. A comparison of transamination
in the unfrozen and the frozen muscle is given in table 2.

Table 2.

Anaerobic iran^auiinaiion in unfrozen and frozen muscles.

Substrates

added

Muscle

before

experi-

Total amino-N
in mg/cc

Amino-N in
glutamic acid
fraction in
mg/cc

A in
glutamic
acid in
mg/cc

Ketonic acid
mg'cc

mcnt

0 min.

30 min.

0 min.

30 min.

0 min.

30 min.

Glutamic acidi

unfrozen

0.68

0.69

0.34

0.26

— 0.84

3.9

4.0

+ pyruvic acid)

frozen

0.58

0.34

0.27

— 0.74

4.1

4.0

Glutamic acid.

unfrozen

0.67

0.68

0.31

0.30

- 0.10

0.2

0.8

n-ketoglutaric J

unfrozen

0.G9

0.7O

0.21

0.23

+ 0.21

3.1

3.2

acid + alanine 1

frozen

0.70

0.17

0.26

+ 0.84

3.0

3.1

«-ketoglutaric
acid ....

frozen

0.15

0.16

0.07

0.08

+ 0.10

2.9

3.0

Blank ....

frozen

0.16

0.15

0.06

—

0.2

0.3

3 g. mnsde + 1.2 cc of 2 per cent EHCO3 + ■water; toial ■v’olume 15 cc.
Sodium arsenitc m/100, bromoacetate 1:5000. Glutamic acid^mg/cc + pyruvic
acid 4 mg/cc. n-ketoglutaric.acid 3 mg/cc + 1 ( + ) — alanine 0.6 mg/cc.

238

GUNNAK IqRBN.

Several experiments, all -with results pointing in the same
direction as the values in table 2, emphasize the conclusion that
the reaction between glutamic acid and pyruvic acid is not con-
siderably influenced by freezing the muscle. On the other hand
the reaction between a-ketoglutaric acid and the NHs-donators
(alanine and peptides) is decidedly stronger after freezing the
muscle. By this procedure the latter reaction is brought up to
the same level as the reaction between glutamic acid and pyruvic
acid. It is most probable that some factor in the reaction between
a-ketoglutaric acid and alanine would be more easily extract
able by prebminary freezing. There was no influence noted on the
total van Slyke nitrogen neither on the ketonic acid concentration
by this procedure. In the following frozen muscles were always
used in reactions between a-ketoglutaric acid and NH 2 -donators.

A comparison was also made between transamination in cattle
muscle suspensions in aerobic and anaerobic conditions. The re-
sults obtained from a series of experiments are given in table 3.

Table 3.

Transamination during anaerobic conditions.

Substrates added

Amino-N in
glutamic acid
fraction in
mg/..

A in
glutamic
acid
fraction

a-ketoglutaric
acid in mg/cc

0 min.

30 min.

in mg/cc

0 min.

30 min.

Glutamic acid - 1 - pyruvic acid|

anaerobic

0.33

0.25

— 0.74

3.9

4.0

aerobic

0.33

0.25

-0.84

4.0

Glutamic acid <

anaerobic

0.38

0.29

- 0.42

0.3

0.2

0.33

0.8

aerobic

0.29

- 0.42

0.8

rt-ketoglutaric acid + alanine |

anaerobic

0.17

0.27

+ 1.05

2.9

3.0

aerobic

0.17

3.0

0.29

- 1 - 1.26

3.0

a-ketoglutaric acid . . . . |

anaerobic

0.18

0.22

+ 0.42

3.1

3.0

aerobic

0.17

0.22

+ 0.62

3.0

2.9

3 g. frozen muscle + 1.2 cc of 2 per cent KHCO 3 ■*" "water; total volume 15 cc.
Sodinm arsenite m/100, bromoacetate 1:5000. Glutamic acid 3 mg/cc. + pyruvic
acid 4 mg/cc. a-ketoglutaric acid -1 3 mg/cc + alanine 3.6 mg/cc.

It is clearly demonstrated that the results gained aerobically
are essentially the same as those gained anaerobically. If anything,
these experiments show that the formation and the disappearance
of glutamic acid are slightly higher under aerobic conditions. The

TRAXSAMINATIOK IN CATIIiE DIAPHRAGM MUSCLE. 239

effect of adding inliibiturs, bromoacetate and sodium-arsenite, ivas
also investigated. These inhibitors vrere recommended by Bratot-
STEiN and Kjutzjiann (1937) in aerobic experiments to separate
the transamination reaction from other lateral reactions nhich
could interfere -with the quantitative results. A typical series of
an experiment is listed in table 4.

Table 4.

The effect of bromoacetate and sodium arseniie on aerobic
transaininaiion.

Substrate added

Inbibitor

Amino-N in
glutamic acid
fraction in
mg/cc.

A in
glntamic
acid in
mg/cc.

Ketonic acids
in mg/cc.

0 min.jso min.

•i

30 min.

Glutamic acid + pyruvic (

0.34

0.26

-0.95

3.9

4.0

acid [

0.34

0.26

-0.84

4.1

4.0

Glutamic acid |

‘h

0.33

0.29

- 0.42

0.2

0.3

0.83

0.29

— 0.42

0.3

0.8

0.09

0.25

+ 1.65

3.0

a-ketoglntaric acid+ alaninet

0.17

0.32

+ 1.57

3.1

O.io

0.14

+ 0.42

3.0

3.1

n-ketoglutaric acid . . . . ■<

0.15

0.19

+ 0.42

3.1

3 g. frozen musde + 1.2 cc. of 2 per cent KHCO3 + water; total volnme 15 cc.
Sodium arsenite m/lOO, bromoacetate 1 ; 5000. G-lntamic acid 3 mg/cc. + pjravic
acid 4 mg/cc. «-ketoglutaric acid 3 mg.'cc + alanine 3.6 mg/cc.

The addition of bromoacetate and sodium arsenite does not in-
fluence the transamination level. If anything, the transamination
mas more effective mhen the inhibitors mere added. Under such
circumstances as these the folloming experiments mere carried
out aerobically mith frozen muscles and in the presence of sodium
arsenite and bromoacetate.

Transamination mitli Peptides.

In morking mith peptides as NHs-donators one of the difficulties
met mith mas the hydrolyzing effect of the muscle peptides. In
a preceding paper (Agren, 1940 c) it mas demonstrated horn at
least tmo di-peptides, glycyl-aminobenzoic acid and valyl-glyoine^
mere but slomly attacked by the peptidases in the cattle muscle.

240

GUNNAK AQRBN.

The experimental conditions •were the same as those demonstrated
above, so as to be optimal in transamination. A typical experi-
ment "with glycyl-aminobenzoic acid is recorded in table 5.

Table 5.

Glutamic acid formation from oL-Jcetoglutaric acid + glycyl-amino-
benzoic acid. Aerobic experiments.

Substrates added

Total amino-N
in mg/cc

Amino-N in glutamic
acid fraction in mg/cc

A in
glutamic
acid
fraction
in mg/cc

0 min.

60 min.

Glycyl-aminobenzoic acid -f
«-ketoglutaric acid . . .

0.54:

0.54

0.12

0.14

0.15

+ 0.32

«-ketoglataric acid ....

0.18

0.07

0.08

+ O.IO

alanine +«-ketoglutaric acid

0.72

0.18

0.26

+ 0.84

3 g. frozen muscle + 1.2 cc of 2 per cent of KHCO 3 + "water; total volume 15 cc.
Sodium arsenite m/lUO, bromoacetate 1 ; 5000. «-ketoglutaric acid 3 mg/cc + glycyl-
aminobenzoic acid 5 mg/cc, alanine 3.6 mg/cc.

T'wo facts were observed when comparing the peptide and the
amino acid reaction in many experiments. The peptide NHa-
donator was slower in its reaction and the transamination did not
proceed to the same extent as "with the amino acid donator.
Only after 60 minutes did the reaction with glycyl-aminobenzoic
acid cease. Yalyl-glycine appeared to be a more easily reacting
peptide, as demonstrated in table 6.

Table 6.

Glutamic acid formation from oi-Tceloglutaric acid and valylglycine.

Aerobic conditions.

Substrates added

Total-N in
mg/cc

Amino-N in
glutamic acid
fraction in
mg/cc

A. in
glutamic
aci4 in
mg/cc

0 min.

15 min.

0 min.’

15 min.

c-ketogluvaric acid -r valylglycine .

0.71

0.72

0.11

0.17

+ 0.69

c-ketoglntaric acid ■ .

0.17

0.08

0.09

+ 0.10

«-ketoglntaric acid + alanine . . .

0.71

0.72

0.18

0.26

+ 0.73

3 g. frozen muscle + 1.2 cc of 2 per cent KHCO 3 + 'water; total volume 15 cc.
Sodium arsenite m/100, bromoacetate 1 : 50{X). a-ketoglutaric acid 3 mg/cc +
valylglycine 7 mg/cc, alanine 3.6 mg/cc.

TRANSAMINATION IN OATTLB DIAPHRAGM MUSCLE. 241

The effectivity of valyl-glycine as NHs-donator is clearly dem-
onstrated. The shorter time of reaction (15 min.) was adapted
in the experiments ivith valyl-glycine, as preliminary experiments
had shoiMi that exceptionally slight hydrolysis of valyl-glycine
could be traced after 30 minutes of incubation. The reaction with
valyl-glycine was of the same order of magnitude as with alanine.

About 30 per cent of added NHa-donators were consumed. The
total %’an Slyke nitrogen values were constant during the experi-
ments, indicating that valyl-glycine was transformed to a-heto-
valeryl-glycine.

Experiments were also carried out to test the reversibility of
the reaction:

valyl-glycine a-ketoglutaric acid a-ketovalerylglycine -f-

-f glutamic acid

Braunstein and Kritzmann reported (1937) that lactic acid
could act as acceptor in transamination experiments. They
suggested a preliminary dehydrogenation of lactic acid to pyruvic
acid. A test was made to discover whether a similar mechanism
existed in cattle muscle and could be used for transamination in
the above-mentioned reaction. The corresponding a-hydroxy-
valeryl-glycine was prepared from a-brom-valeryl-glycine. In a
similar way a-hydroxypropionyl-glycylglycine was obtained. A
few results "with these substrates are given in table 7.

Table 7.

Transamination tuiih a.-hydroxyacids in peptide linTcages as acceptors.

Substrates added

Total amino-N
in mg/cc

Amino-N in
glntamic acid
fraction in
mg/cc

A in
glntamic
acid in
mg/cc

Ketonic acids
in mg/cc

0 min.

30 min.

0 min.

30 min.

0 min.

30 min.

Glntamic acid+pymvic
acid

0.50

0.61

0.83

0.26

— 0.68

4.1

Glntamic acid ....

0.61

0.29

0.2

0.8

Glntamic acid + lactic
acid

0.49

0.28

0.27

-O.IO

0.8

Glntamic acid + a-byd-
roxyglycylglycine .

0.46

0.28

0.66

— 0.82

0.8

0.4

Glntamic acid+n-hyd-
roxyvalerylglycino .

0.4G

0.46

0.28

0.26

-0.21

0.4

3 g. frozen mnsde + 1.2 cc of 2 per cent K.HCO3 + water; total volnme 15 cc.
Sodinm arsenxte m/lOO, bromoacetate 1:5000. Glutamic acid 3 mg/cc+pjrnvic
acid 4 mg/cc, lactylglycylglycine 7.6 mg/cc, lactic acid 3.6 mg/cc. n-hydroxyval-
erylglycine 7 mg/cc.

242

GUNNAR IgREN,

There seemed to be a tendency towards a-hydroxyacids in pep-
tide linkage witli an amino acid functioning as ITHz-acceptors.
Experiments witli the corresponding a-ketonic acids met with
difficulties in the preparation. The problem was attacked in an
indirect way. a-ketoglutaric acid and the NHz-donator (amino
acid or peptide) were brought to reaction in the muscle suspension.
Then the reaction was forced back again by adding a surplus of
glutamic acid. The result of a typical series is given in table 8.

Table 8.

The reversibility of transamination with amino acids and peptides.

Aerobic conditions.

Substrates added

Total amino-N
in mg/cc

Amino-N glut-
amic acid frac-
tion in mg/cc

A in glutamic
acid in
mg/cc

Ketonic acids
in' mg/cc

min.

«-ketoglutaric acid
+ alanine = Cojo .

0.46

0.18

0.17

-b 0.42

3.0

3.1

a-ketoglutaric acid
+ alanine. After 20
min. stopped im-
mediately on tbe
addition of glut-
amic acid = GIA20

0.46

0.68

0.39

3.1

As GIA20 but stopp-
ed 10 min. after
adding glutamic
acid = GIA30 . .

0.46

0.64

0.39

0.37

— 0.21

3.2

a-ketoglutaric acid

0.15

—

0.14

0.06

—

0.07

—

+ 0.10

3.1

—

a-ketoglntaric acid
-f valylglycine . .

0.71

0.72

—

0.16

0.27

—

+ 1.12

3.2

3.1

—

As GlAjo but valyl-
glycine substitut-
ing alanine . . .

0.90

0.42

3.2

As GlA3r, but valyl-
glycine substitut-
ing alanine . . .

0.91

0.37

-0.68

3.1

Glutamic acid 2
mg/cc

O.Sl

0.81

0.20

0.19

-0.10

0.3

0.2

3 g. frozen muscle + 1.2 cc of 2 per cent KHCO3 + -water; total volume 15 cc.
Sodium arsenite m/100, bromoacetate 1 : 5000. c-ketoglutaric acid 3 mg/cc +
alanine" 1.8 mg/cc. After 20 min. incubation addition of glutamic acid to a
concentration of 2 mg/cc. Valylglycine 7 mg/cc.

TRAXSAMIKATTON KT CATTLE DIAPHRAGM JIDSCLE. 24S

The table demonstrates the possibilities of influencing the
stationary level in the two reversible reactions:
a-ketoglutaric acid 4- alanine glutamic acid -j- pyruvic acid
a-ketoglutaric acid valyl-glyciue ■jrt glutamic acid + a-keto-

valeryl-glycine

Starting from left to right, the end point obtained can be
forced back a little to the left by adding glutamic acid. Thus it
appears reasonable to assume that a-ketovaleryl-glycine can
function ns NHs-acceptor to form a dipeptide. The assumption is
emphasized by the constancy of the total van Slyke nitrogen
values.

The difference in reactmty of the peptides and amino acids
raised the question of which substrates could function together
with a-ketoglutaric acid to form glutamic acid. In the first scries
of experiments, quoted above, a l(-}-)'alanine preparation was
used. Being without this preparation for a time, a d, 1-alanine,
in a concentration tnnee as high, was added. The }aeld of glutamic
acid in the latter series was much higher than expected. Conse-
quently, the optical isomers and the racemic compounds of a few
amino acids were examined. The results of these experiments are
listed in table 9.

Table 0.

Glutamic acid formation from a.-}cetogluiaric acid and different

amino acids.

Amino acids added

Number of
experiment

Amino-N in
glutamic ncid
fraction in mg/cc

A in
glutamic
ncid in

0 min.

30 min.

mg/cc

1 (+) alanine

0.11

0.17

+ 0.68

d, 1 alanine

0.17

0.27

+ 1.05

1 (+) alanine

0.14

0.22

+ 0.84

d (—) alanine

0.14

0.17

+ 0 .S 2

d, 1 alanine

0.12

0.21

+ 0.95

1 (+) valine

0.07

0.11

+ 0.42

d (— ) valine

0.09

0.11

+ 0.21

d, 1 valine .

0.08

0.12

+ 0.42

Experiments Tvitb tbe identical numbers carried out •srith muscle suspensions
&om the same muscle. The values corrected for the spontaneous reaction of
a-ketoglntaric acid. 3 g. frozen muscle -h 1.2 cc of 2 per cent KHCO 3 - 1 - water;
total volume 15 cc. Sodium arsenite m/100, bromoacetate 1:5000. «-ketoglnt-
aric acid 3 mg/cc, 1 {+) and d (— )-alamne 1.8 mg/cc. d, 1-aIamine 3.6 mg/cc.
1 (+) — and d (— )-valine 2 mg/cc, d, 1-valine 4 mg/cc.

244

GUNNAR IgREN.

A number of experiments 'were made, all varying in tbe same
■way in the results. Cattle muscle seems to have a tendency to-
wards forming glutamic acid from the unnatural forms of amino
acids, but this tendency varies in different muscles.

Discussion. It has been proved in the first section of the pres-
ent paper that cattle diaphragm muscle can be used as a reliable
resource for enzymes, when wprldng at transamination experi-
ments. This is of some practical interest, as this muscle can be
obtained in great amounts immediately after the death of the
animals. Braunstein’s and Kritzmann’s results (1937) regarding
the formation of glutamic acid from a-ketoglutaric acid and ala-
nine have been corroborated. On the other hand a discrepancy
was discovered in the reaction of ketonic acids. The Russian
workers found that ketonic acids disappeared during the trans-
amination. This result was interpreted as a transformation of
a-ketoglutaric acid to succinic acid. It is evident that this lateral
reaction, including the decarboxylation of a-ketoglutaric acid,
is not as apparent in cattle muscle as in pigeon breast muscle.
There may be some variation in the activity or concentration of
the decarboxylating system, according to how the requirement
and metabolism in carbohydrate varies. On the other hand this
variation may be a secondary one, originating in the release of
otherwise structurally coimected enzymes.

It is of interest to observe that a-keto groups in peptide link-
ages to amino acids can function as acceptors of amino groups and
form dipeptides. The part played by such a mechanism in protein
metabolism -will not be discussed in this connection, but it has
been remarked before that such compounds could be formed in
a reaction between ketoaldehydes and amino acids. The varying
results in the experiments -ndth the optical isomers af amino acids
may be of interest in the light of Kogl’s results (1939). An in-
di-vidually varying actmty in the transformation of unnatural
forms of amino acids is a possibility worth paying attention to.

The author "mshes to express his gratitude to Professor K. Lm-
derstrom-Lang for extending the facilities of the Carlsberg Labor-
atory and for his untiring interest in the work.

TKANSAMINATION IN CATTLE BIAPHRAGM SlUSCLE.

245

Summary.

1. Transamination has been studied in cattle diaphragm
muscle. The reaction proved, with slight discrepancies, the well
kno\s-n characteristic properties.

2. The reaction between a-ketoglutaric acid and amino acids
or peptides was facilitated by preliminarily freezing the minced
muscle. A disappearance of ketonic acids could not be demon-
strated. The transamination was not inhibited by bromoacetate
or sodium arsenite.

3. Glycyl-aminobenzoic acid and valyl-glycine coiild be used
as amino group donators. In the opposite way glutamic acid
apparently reacted with a-hydroxy and a-ketonic acids in peptide
chains.

4. Muscles from different animals displayed some variability
in the reaction between c-ketoglutaric acid and unnatural amino
acids and racemic derivates.

References.

Braukstein, a. E., and M. G. Kritzmann, Enzymologia 1937. 2 . 129.
Braunstein, a. E., Ibidem 1939. 7. 25.

Conway, E. J., and A. Byrne, Biochem. J. 1933. 27 . 419.
Fromageot, C., and P. DESNimnLEs, Biochem. Z. 1935. 279 . 174.
Jones, D., and 0. Moeller, J. biol. Chern. 1928. 79 . 429.
Kritzmann, M. G., Enzymologia 1938. 5. 44.

BIogl, F., and H. Erxleben, Hoppe-Seyl. Z. 1939. 268 . 57.
Needham, D. M., Biochem. J. 1930. 24 . 208.

Neuherg, C., and G. Ringer, Biochem, Z. 1915 — ^16. 71 — 72 . 228.
Agren, G., and E. Hammarsten, 0. R. Lab. Carlsberg, Ser. chim.
1938. 22 . 25.

Agren, G., H. Hammarsten, and K. G. Rosdahl, J. biol. Chem. 1939.
127 . 541.

Agren, G., C. R. Lab. Carlsberg Ser. chim. 1940 a. 23 , 173.

— , This Journal 1940 b. 1 . 105.

— , Ibidem 1940 c. 1 . 119.

From the Pharmacological Department of the Karolinska Institute,

Stockholm.

The Absorption of Ethyl Alcohol from
the Gastro-Intestinal Tract as
a Diffusion Process.^

SVEN M. BERGG-REN and LEONARD GOLDBERG.
CWith 4 figures in the text.)

Introdnction.

Previous investigations on the distribution of alcohol in the
system indicate that this process must follow certain laws (Gr:^-
HANT 1895, Carpenter 1929, Widmark 1930 and Nicloux
1934), and a number of investigators have been able to prove
that its passage to various fluids in the body takes place by
diffusion (urine: Widmark 1916, milk: Glow 1923, cerebrospinal
fluid: Abramson and Linde 1930, gall: Niclodx 1931, saliva:
Linde 1932, sweat: Nyman and Palmlov 1936, aqueous humour:
Berggren). There is therefore reason to assume that the same
process applies to the absorption from the gastro-intestinal
tract. There are, however, very few references to support this
assumption clearly.

Direct examinations of the contents of the stomach and the
intestines show that all parts of the alimentary canal are cap-
able of absorbing alcohol (Tappeiner 1880, Nemser 1907, Hanz-
iiK and Collins 1913, Lucas 1930, Carpenter 1937). In some
cases the alcohol percentage has been determined in the contents
of the stomach or the intestines running out of fistulae (Kautsch
1898, Lonnquist 1906, Nemser 1907), in some instances by

* Received 6 September 1940.

ABSOIIPTION OP ETHYL ALCOHOL.

247

direct determination of the contents when withdrawn from the
stomach after ligating the pylorus or from isolated loops of the
intestines (Tappeiner 1880, Segall 1888, Brandl 1893, Ha-
NEBORG 1921, Edkixs and Murray 1926) and lastly by analys-
ing the contents after killing different animals at different intervals
(VoLTz and Dietrich 1915, Cori, Villiaume and Cori 1930,
Le Breton 1936, Harger, Hulpieu and Lamb 1937). These
investigations are, how’ever, incomplete to a certain extent and
do not give absolutely comparable results, in consequence of
which it has not been possible to obtain a true picture of the
changes that the alcohol percentage undergoes in the contents
of the stomach and the intestines; generally spealcing, however,
it seems as if the alcohol is more quickly absorbed during the
first hour of the experiment and also to a fair extent already
in the stomach.

Several investigators have also endeavoured to come to some
conclusions concerning the course of absorption from the point
of the blood alcohol curve;

When alcohol is ingested on an empty stomach absorption
seems to take place comparatively quickly and to be completed
in 1 or 2 hours (Mellanby 1919, Widmark 1932, Bernhard
and Goldberg 1935), others maintain that it takes 5 — 6 hours
(Haggard and Greenberg 1935). A stronger percentage of the
alcohol ingested seems to bring about a more rapid absorption
(Mellanby 1919, JIiles 1922), though Handwerk (1927) and
Tuovinen (1930) contradict this, while other investigators have
not been able to find any difference (Widmark 1932, Schmidt
1937). Absorption seems to take place more quickly in those
habituated to alcohol than in normal indi^dduals (humans:
Sohweisheimer 1913, Jungmichel 1933, Schmidt 1934, Fle-
ming and Stotz 1935, Bernhard and Goldberg 1935, animals:
Yoltz and Dietrich 1915, Faure and Loew^ 1923, Newman
and Card 1937), something that has not been confirmed by some
investigators (humans; Miles 1922, Matossi 1931, Graf and
Flake 1932, animals: Gettler and Freireich 1935). — Si-
multaneous ingestion of food delays absorption, and some authors
look upon the constituent parts of the food as playing a certain
role (Mellanby 1919, Southgate 1925, Widmark 1934, Neymark
and Widmark 1936, Schmidt 1937), other investigators take the
caloric percentage into consideration (Jungmichel 1933), while
Elbel and Lieck (1936) attach importance to the quantity.

248

SVEN M. BBRGQREN AND LEONARD GOLDBERG.

Prom the literature on the absorption of alcohol it may be
summarily concluded that ethyl alcohol is obviously absorbed
with gradually diminished rapidity according to the period of
time from the beginning of the absorption. The greater quan-
tity of alcohol is absorbed during the first hour, unless large
doses have been given, but in particxilar cases considerable un-
absorbed quantities may be found even after several hours. The
concentration of the alcohol solution has been assigned different
significance in different experiments. The simultaneous ingestion
of food produces a delay in the absorption, but to what extent
this may be regarded as having a diluting effect has, according
to our opinion, hitherto received too little consideration.

Own Inyestigations.

Mathematical Analysis.

The object of this investigation was to follow the change in
the concentration of ethyl alcohol in the contents of the stom-
ach during absorption by means of direct determinations and
to see whether there is any criterion for the assumption that the
absorption is a diffusion process. Before describing thfe methods
and their results, a general analysis is given of the factors in-
fluencing the passage of alcohol from the gastro-intestinal tract
to the system. For this analysis it must be assumed that alcohol
diffuses freely through all the tissues of the body and that the
concentration when equilibrium has been established is fixed
by the percentage of water in the different tissues.

Assuming a pure diffusion, the factors establishing the disappear-
ance of alcohol from the contents of the stomach are as follows:

The concentration of the ingested alcoholic solution.

The motility of the gastric wall (the mixing up of the contents).

The permeability of the gastric mucous membrane.

The number of capillaries in the mucous membrane.

The blood flow through the mucous membrane.

According to Pick’s law of diffusion, the rapidity of diffusion is
directly proportional to the difference of the concentration between
the two media (in this case the contents of the stomach and the blood).
On account of the alcohol percentage in the blood during the greater
part of the absorption being slight in comparison with the percentage
of alcohol in the contents of the stomach, the former may be ignored.
Expressed in mathematical form, the process of diffusion may be signified

_|?.V=a.c (1)

ABSORPTION OF ETHYL ALCOHOL.

249

Ttus ^ gives tlie change in concentration pr time unit ( — sign

gives decrease in concentration),

V = quantity of alcoholic solution,
a = diffusion constant,
c = concentration of alcoholic solution.

Development of above formula gives

Integration

In c = |;yr • t + k j

c == e ' • e~'';

When t = 0; c = e—

c = Co; e- k = Co,

--•t

Thus c = Co • e (2)

The quantity of alcohol in the tissues of the body on a given occa-
sion (t) may then be defined as follows;

V.Co — V-Ct, — /3-r.t-p

P = rate of disappearance of alcohol from the blood in promille
pr min,

r = relation between percentage of alcohol in the body and in
the blood,

t = time in min,

p = body weight in kg.

After replacing of Ct according to formula (2)

V-Co — V-Co-e“'^‘‘— /3-r.t.p (3)

If V and Co in formula (3) are allowed to vary in an opposite direc-
tion, i. e. to allow the absolute quantity of alcohol ingested to be iden-
tical but to let the quantity of the solution ingested and the concen-
tration vary, it will obviously bring about changes in the total quan-
tities of alcohol existing in the system at fixed times. An increase
in the concentration (Cq) of the solution ingested and a corresponding

decrease in the volume V will produce an increase in the factor • t,

thus an increase in the whole expression. This implies that the total
remaining quantity of alcohol in the system with the same absolute
amount ingested will for a certain time be greater as the concentra-
tion of the solution ingested becomes higher. By mailing use of for-
mula (3) for the quantity of alcohol in the system, it has been calcu-
lated (fig. 1) what values of alcohol percentage can be expected in

^50

SVEN M. BERQGKEN AND LEONARD GOLDBERG.

tlie body, -wbeii 1 g alcohol pr kg body weight is ingested of solutions
of 10 % (I) and 50 % (II) respectively. In addition to this a curve
has been calculated with a solution of 10 % with a double value of
the diffusion constant (III). Complete diffusion equilibrium between
the different parts of the body has been considered to exist.

Kg. 1.

Alcohol percentage in system after ingestion of 1 g pr kg body weight.

I Solution 10 %. Diffusion constant 0.02

n » 60 % » » 0.02

m » 10 % » » 0.04

Bate of disajjpearance of alcohol from blood 0.0020 pr min.

From the curves in fig. 1 it will be seen what variations can be ob-
tained in the size as well as in the time of the maximum of the alcohol
percentage, vital factors for the appearance of intoxication symptoms.
The greater the concentration of the alcohol ingested and the less the
dilution taking place in the stomach owing to liquid ingested (even
in the form of food), the greater will be the value for the alcohol con-
centration and the sooner will the maximum appear.

The influence of variations in the blood flow is seen from the fol-
lowing:

The blood flowing through the mucous membrane of the stomach
absorbs a certain quantity of alcohol, which varies with the diffusion
constant for the mucous membrane and the concentration fall between
the contents of the stomach and the blood. While without running

ABSORPTION OF ETHYL ALCOHOL.

251

the risk of any miscalculation it may be said that the alcohol percen-
tage of the arteriolar blood is small as compared vrith that of the con-
tents of the stomach, such an assumption is only likely for the blood
after passage through the capillaries of the mucous membrane ■where
diffusion takes place, if the flo'w of blood is rapid. On the other hand,
if the blood flo'w is slow, a considerable increase in concentration must
take place, and the difference in the concentration between the con-
tents of the stomach and the blood must be decreased, which -will con-
sequently delay the diffusion. .

The curve representing the rise of the alcohol percentage in the blood
during its passage through the mucous membrane is also an expo-
nential function, which asymptotically approaches the alcohol per-
centage of the contents of the stomach, which may be regarded as
constant during the short time the same quantity of blood remains
in the mucous membrane.

If C5 = blood alcohol percentage,

Cv — alcohol percentage of contents of stomach,
a = diffusion constant,

B = blood flow through mucous membrane pr time unit,
t = time (t = 0 means the time when the blood passes from
arterioli to capillaries),

then the quantity of alcohol diffusing from the stomach to the blood
may be signified according to Pick’s law thus

^.B = a ((v-Cb)

which after intogratidn gives

--•t

Cb = — C^-- e B (4)

If the quantity of blood (B) flo'wing through the mucous membrane
pr time unit is constant and passes "with constant rapidity (which im-
plies that the time for the passage of the capillaries is constant) and
the alcohol percentage of the onflow of blood is minimal, the alcohol
percentage in the blood of the mucous membrane (Cb) must approach
a value, which stands in a constant relation to the alcohol percentage
in the contents of the stomach. If this relation is defined by K and
the above curve is calculated for the concentration K • Cy, which gives
the alcohol percentage of the blood of the mucous membrane after
its passage through the capillaries, the following is obtained

K'Ct

Cy — Ct • e ®

K==l — e (5)

From this it will be seen that K increases with increased rapidity
of diffusion through the mucous membrane and decreases "with in-
creased blood flow.' These relations are evident from the curves (IT)
and (IQ) in fig. 2.

18 — i01S23. Acta phys. Scandinav. VoL I.

252

SVEN M. BERGGREN AND LEONARD GOLDBERG.

During the passage througli the capillaries of the mucous membrane
the value of the blood alcohol percentage (Cb) increases from approx-
imately 0 to K • c,,. The average for the values of the blood alcohol
in the different parts of the capillaries can be calculated from formula
(4). This calculation is shown in fig. 2, where the time for the passage

Fig. 2.

Alcohol percentage in blood during passage through capillaries
of gastric mucous membrane.

I Alcohol percentage in contents of stomach

n » » » blood. Diffusion factor 0.3

in » » » » » » 0.1

OB Time for passage of capillaries.

Alcohol percentage of contents of stomach
K • C.J. » » » blood after passage through capillaries,

k ■ Average value of alcohol percentage of blood during passage (see test).

through the capillaries in the mucous membrane of the stomach has
been chosen to be 4 sec. The value of the alcohol percentage in the
blood after the passage through the capillaries is according to the
above assumption K • c^. The area OAB is determined by definite
integration of the function (4). This area corresponds to a rectangle,
the height of which represents the average of the alcohol percentage
during the passage through the capillaries. This value has been called
k* c^. If the blood flow is constant, k like K will be a constant. The
relation between k and K can be calculated from formulae (4) and
(5) and is

k =

1-f

In a — K)

( 6 )

From formula (6) it is evident that k varies in a parallel manner
with K.

ABSORPTION OP ETHYL ALCOHOL.

253

The concentration difference deter minin g the alcohol diffusion from
the stomach to the blood — the blood alcohol percentage taken into
consideration — ■will ob'nously vary between the values Cy- and
Cy — K • Cy round an average value, which according to the above can
be expressed Cy — k* Cy. The total alcohol percentage in the contents
of the stomach -will then, according to Fick’s law, fall as follo'ws:

^•V = a(Cy — k-Cy) (7)

Development of this gives

Integration

dcy a • dt

(1-k)

In Cy =

a-t (1 — k)
V

( 8 )

Cy = Gy^ . e

(9)

With the exception of the exponents formula (2) and formula (9)
coincide completely. Taking into consideration the alcohol percentage
in the blood thus makes no essential difference to the mathematical

expression for the process of diffusion. The exponent of e = — ^

can be empirically determined from the value of the concentration,
and a constancy of the thus calculated exponent during one and the
same experiment indicates that the process has been one of diffusion.
On the other hand, the determination of the actual diffusion constant
(a) involves considerable difficulties, since both the concentration in
the blood of the mucous membrane reached during the passage and
the quantity, of the blood flow are not accessible for experimental
determination.

Methods.

I. In one series cats have been employed after they have
been kept starving for 12 hrs. The animals were anaesthetized
with Pernoctone, 0.6 — 0.7 ml pr kg subcutaneously. In order to
remove any likely remains of food and secretion the stomach
was flushed several times with Einger solution by means of a
small rubber 'tube. Through the tube was ingested an alcoholic
solution of 5 — 10 per cent by volume in a quantity of approxi-
mately 10 ml pr kg. Samples (0.1 — 0.2 ml) were taken of the
contents of the stomach by means of the stomach tube at regu-
lar in-tervals, and after suitable dilution the alcoholic concen-

254

SVEN M. BBRQGREN AND LEONARD GOLDBERG.

tration was analysed according to Widjiark’s micrometliod
(1932) witli the modifications introduced hy Linde (1932).
The quantitative contents of the stomach were withdrawn and
measured several times during each experiment, which as a rule
lasted for 1 or 2 hours.

In some cases, too, when narcosis had set in, the pylorus was
ligated, and particular care was taken not to obstruct the vessels
r unni ng along the pylorus, as in that case a decrease in the hlood
flow through the gastric wall would in all probability have en-
sued. After the experiments — : several experiments were made
on the same animal — the animal was killed and the stomach
closely examined in order to control the ligation.

During the short time the experiment lasted constant circu-
lation conditions in the mucous membrane of the stomach can
be reckoned with, apart from momentary changes which took
place when removing the contents quantitatively. In conse-
quence of the inability of the stomach to absorb water (see be-
low), a definite constancy of the volume of the alcoholic solution
ingested was obtained in the experiments with ligated pylori.

II. In a second series a number of experiments were made
on human subjects. 300 ml of an alcoholic solution of 5 per cent
by volume were given through a thin stomach tube and samples
(0.5 — 1.0 ml) for determination of alcohol were withdrawn every
5 — 10 min during 1 — 2 hours.

Based on the values of the alcohol concentration obtained, the

a (1 — k) . ,

factor — of the exponent to e in formula (9) was cal-

culated. This factor, which comprises the diffusion constant of
the mucous membrane for alcohol (a), the volume of the alco-
holic solution (V) and the coefficient k, which is a function of
the diffusion constant (a) and the blood flow (B) — cp. formulae
(6) and (6) — has been termed the diffusion factor. If 'the
passage of the alcohol from the contents of the stomach to
the blood is a simple diffusion process, and the factors of the
exponent are constant (see p. 252) the diffusion factor will also
be a constant.

In order to equalize the accidental error in the determination
of the alcohol percentage, the factor has been calculated con-
tinuously between two consecutive determinations (ci and c,)
in one of the following ways:

ABSORPTIOK OF ETHYL ALCOHOL,

255

a) By calculating tte decrease in the alcohol percentage pr

min for every interval hetveen two determinations (— This

\dt/

decrease being an expression for the rapidity of the fall in the
curve, i. e. its differential coefficient, is inserted in formula (8),
where c is approximated to the average between the values of

the two determinations: ^ — = c„.

b) By a direct insertion of the values of the determinations
(ci and c.) and the time interval (t) in formula (9).

Por values of the diffusion factor up to 0.035, both methods
give practically the same results, the difference being 0 — 2 %,
in which case the first method (a) is considerably more simple
and also has been generally used in the following experiments.
"When the curve falls at a greater rate, u c. for values of the dif-
fusion factor above 0.035, the difference will be greater. Eor
values of 0.050 — 0.120 the difference between the two methods
used is 2 — 8 %, method (a) giving the lower values, for values
round about 0.170 the difference is about 25 %. This is due to
the fact that with the same time intervals between the deter-
minations there are relatively fewer ones at a more rapid fall
of the curve, in consequence of which the approximation accord-
ing to method (a) will be charged with a greater error.

Owing to the time intervals between two determinations ge-
nerally being fairly short — 6 — 10 min — , the difference in con-
centration between two consecutive determinations will also be
relatively slight. If, as is the case in the majority of the deter-
minations, this difference amounts to approximately 20 % of
the existing concentration, and the error for the alcohol deter-
mination is i; 5 % (including dilution and determination er-
rors), the maximum error for the difference between two deter-
minations amounts theoretically to 50 %. Practically the stan-
dard deviation (a) lay between 14 and 35 %. the average being
± 23 %, and the standard error of the mean (e) between 6 and
20 %, the average being ± 12 %.

256

SVEN M. BBBGGBEN AND IiEONABD GOLDBEBG.

Alcohol percentage in contents of stomach.

Standard deviation ± 29 %, standard error of the mean ± 11 %.

It will be seen from fig. 3 bow tbe determinations of tbe alco-
bol percentage in tbe contents of tbe stomach are grouped round
tbe theoretically calculated exponential curve during tbe course
of absorption, tbe standard deviation in this case being ± %j
tbe standard error of tbe mean i 11 %.

Besnlts.

A. Cats.

Tbe following experiment on a cat may be given as a typical
illustration of tbe course of tbe alcohol percentage in tbe con-

ABSORPTION OP ETHYL ALCOHOL.

257

tents of the stomach, the pylorus being intact; the alcohol con-
centration was obtained through determination of samples with-
drawn at regular intervals.

6. 7. — 38. Exp. 13. Cat IX. 3.G kg. Pernoctone 0.6 ml pr
kg. Pylorus intact. 30 ml of an alcoholic solution of 6 per cent
by volume by stomach tube 16^*45'. Contents of stomach,
withdrawn quantitatively every 10 min, vary between 29 and
30 ml.

Time

min.

Ale.

cone.

®/oo

min.

Diff. fact,
dc

method (a)
I

Log c

Diff. fact,
method (b)

Diffe-

rence

n-i

29.0

1.4624

22.6

6.4

0.64

25.8

0.0248

1.3641

0.0249

— O.OOOl

19.6

3.1

0.81

21.06

0.0147

1.2900

0.0148

— O.OOOl

15.2

4.8

0.48

17.86

0.0248

1.1818

0.0249

— O.OOOl

12.6

2.6

0.26

13.9

0.0187

1.1004

0.0187

±0

10.2

2.4

0.24

11.4

0.0211

1.0086

0.0212

— O.OOOl

8.96

1.26

0.126

9.68

0.0181

0.9618

0.0181

±0

DiffQBion factor : average 0.020 ± 0.002
Alcohol diBoppearing from stomach after 30 min. ; 47.6 %
> > > » > 60 min.: 69

Prom this experiment it is evident that physiologically the
pylorus can remain closed as long as one hour, during which
time no change seems to take place in the volume of the
solution ingested. After 30 min about 48 % and after 60 min
69 % of the alcohol ingested, when given in a solution of
6 per cent by volume, had disappeared. Finally the example
illustrates the constancy of the diffusion factor within the limits
of error, the standard deviation being 0.0046, which is ± 23 %
of the average, and the standard error of the mean being 0.002,
corresponding to 10 % of the average. The difference between
the methods (a) and (b) is less than a quarter of a per cent and
can be neglected as long as the diffusion factor does not exceed
0.036.

Another experiment on the same animal after ligating pylorus
is given below.

258

SVEN M. BERGGEEN AND LEONARD GOLDBERG.

11. 7. 38. Exp. 14. Cat. IX. 3.6 kg. Pernoctone O.Gmlprkg
subcut. 30 ml of an alcoholic solution of 5 per cent by volume
by stomach tube 10^ 11'. Pylorus ligated. Contents of stomach,
withdfa-wn quantitativdy every 10 min, vary between 29 and
30 ml.

Time

min.

Ale.

cone.

®/oo

min.

°/oo

Log c

Diff. fact.
method(b]
11

Difference

n— I

27.6

1.4409

24.7

2.9

0.68

26.15

0.0222

.1.3927

0.0222

i 0

22.0

2.7

0.64

23.36

0.0231

1.3424

0.0282

— O.oooi

18.G

3.4

0.34

20.3

0.0168

1.2696

0.0168

± 0

14.9

3.7

0.37

16.76

0.0221

1.1732

0.0222

- O.oooi

12.1

2.8

0.28

13.5

0.0207

1.0828

0.0208

±0

■ 10.1

2.0

0.2 0

11.1

0.0180

1.0043

0.0181

— O.oooi

6.8

3.S

0.13

8.45

0.0164

0.8826

0.0168

— 0.0004

Diffusion factor : average 0.020 + 0.001
Alcotol absorbed from stomach after 30 min.: 46 %
> j > > >60 min. : 69 %

These two experiments, performed on the same animal, fully
agree (exp. 13 and 14). Here 46 % of the alcohol ingested
disappeared from the stomach after 30 min and 69 % after 60
min. As the volume of the solution had been constant during
the whole time — water not being absorbed from the stomach —
the disappearance of alcohol must be attributed to an absorp-
tion. The diffusion constant had the same value as with intact
pylorus; 0.020 ^0.001, the standard deviation being relatively
smaller: ± 0.0028, that is 14 % of the average, and the standard
error of the mean ± 0.001, which is 5 % of the average. Ho
essential difference could be seen between methods (a) and ,(b)
when calculating the diffusion factor, and in the following' the
diffusion factor will be calculated according to method (a).

The ingestion of alcohol in a solution of 16 — 17 % shows
similar conditions, as is illustrated by the following example.

ABSORPTION OF ETHTL ALCOHOL.

259

27. 8. 38. Exp. 19. Cat XI. 2.91 kg. Pernoctone 0.7 ml pr
kg subcut. 30 ml of an alcobobc solution of 16.7 per cent by
volume by stomacb tube 12^^ 30'. Pylorus ligated.

Time

min.

Ale. cone.

7oo

®/oo

min.

Diff. fact.

86.3

80.8

5.6

1.1

83.6

0.013

73.0

7.8

1.5G

76.9

0.020

66.0

7.0

1.4

69.5

0.020

56.7

9.3

1.8G

61.4

0.030

52.5

4.2

0.84

54.6

0.015

40.1

12.4

1.24

46.3

O.027

30.0

10.1

1.01

35.1

0.029

26.2

3.8

0.38

28.1

0.014

85 »

13.4

12.8

0.43

20.0

0.022

115

10.0

3.4

0.11

11.7

O.OlO

Diffnsion factor : average 0,020 + 0.002
Alcohol absorbed from stomach after 30 min. : 46 3?
> > > > > 60 min. : 72 %

} j > > > 90 min. : 93 %

■ The diffusion factor is in this experiment of the same magni-
■tude as when alcohol in a solution of 5 per cent by volume ^
given, which also supports the assumption that absorption is
;a diffusion process. As in some other cases, another fact appears
•in this example. The diffusion factor, which has on the whole
been constant, —0.021 ± 0.002 — shows a certain tendency
•to decrease towards the end of the experiment and its last value
is as low as 0.010, a figure which lies outside the normal limt
of variation. The most likely reason for this is a deterioration
of the blood flow in the mucous membrane, i. e. a decrease of
B [see formulae (4) and (6)] and consequently in the whole of
the diffusion factor, or else a decrease in the mobility of the

gastric wall with incomplete mixing. _

When the stomach is emptied during the experiment, the
alcohol concentration in the contents of the stomach takes an-
other course. A typical example of this in a case of intact

260 SVEN M. BBRGGRBN AND LEONARD GOLDBERG.

pylorus, "wliere a more or less continuous decrease in the
volume of the solution ingested has taken place, is illustrated
by exp. 6,

9. 6. 38. Exp. 6. Cat VII. 4.0 kg. Pernoctone 0.7 ml pr kg
subcut. 42 ml of an alcoholic solution of 4.8 per cent by volume
by stomach tube 9^* 56'. Pylorus intact.

Time

min.

Volume of
contents of
stcmacli
ml

Ale. cone.
®/oo

®/oo

min.

“/oo

Diff. fact.

31.6

27.7

3.8

29.6

0.026

20.3

7.4

24.0

0.031

14.9

5.4

17.6

0.081

18.5

9.8

5.1

12.4

0.041

14.6

6.46

3.86

8.1

0.028

3.4

3.05

4.9

0.041

10.6

2.3

1.1

2.85

0.026

Diffusion factor : average 0.032 ± 0.008
Alcohol disappearing after 30 min. : 61 %

> > » 60 min. : 86 %

In this case the diffusion factor was 0.032 ± 0.003; after 30
min 61 %, after 60 min 86 % of the alcohol ingested had dis-
appeared. Another experiment on the same animal — also with
an intact pylorus — gave a diffusion factor of 0.039± 0.011;
77 % of the alcohol ingested had disappeared after 30 mm (exp.
7, table 1). These values are to be compared with the results
of experiments on the same animal after ligating pylorus (exp.
8 and 9, table 2). After ligation the diffusion factor was 0,017
and 0.021 respectively, as agamst 0.032 and 0.039 before; after
30 min an average of 38 % of the alcohol ingested had been ab-
sorbed agamst an average of 69 % when the pylorus was left
intact. The diffusion factor had thus increased with 84 %, cor-
responding to an increase in the amount of alcohol absorbed
after 30 min of 82 % The increase of the diffusion factor in the
case of non-ligated pylorus will thus be caused by the decrease
of the volume (V) of the solution ingested, which fully agrees

ABSORPIIOK OF ETHYL ALCOHOL.

261

Tvitli the fornuilae adopted [cp. formula (9), where V is to be
found in the denominator of the diffusion factor]. The difference
between the diffusion factors in these experiments before and
after ligtiting pylorus is statistically significant. — When no
change in the volume of the solution ingested takes place, the
diffusion factor becomes constant, as does also the amount of
alcohol absorbed (cp. c.xp. 12 and 13, table 1 and e.xp, 14 and 15,
table 2).

A survey of the experiments with intact pylori is given in table
1. The strength of the alcoholic solution was 5 per cent by vol-
ume, c:a 10 ml solution was given pr kg body weight.

Table 1.

Surteii of experitnents wUh iniad pylori on cat.

Exp.

Ale.
cone.
Per cent
by vol-
ume

Per cent alcohol
absorbed after

Diflusion

factor

Volume of
contents of
stomach
ml

32.f.

68.5

O.Oll ± O.OOl

Constant

3S..n

.68

—

0.015 ± 0.003

Decreasing

63.5

—

O.OlO ± 0.004

—

0.032 ± 0.003

—

0.039 + 0.011

■rai

—

0.014 ± 0.002

Constant

—

O.OIG ± O.OOl

Decreasing

61.5

—

0.024 ± 0.004

Constant

47.5

—

0.020 ± 0.002

DilTnsion factor vitb intact pylorus ; average 0,021 + 0,OOS (9 exp.)
Alcoliol disappearing after 30 min. : 48 (C exp.)

» > j GO min. : 67 »

The average of the diffusion factors, the volume of the solution
ingested being constant, was 0.017 dz 0.003, the average with
decreasing volume 0.024 J:; 0.006, the difference not being sig-
nificant as the number of experiments was too small to allow
of statistics.

Below is given a similar survey of cases with ligated pylori
(table 2).

262

SVEN M. BERQGRBN AND LEONARD GOLDBERG.

Table 2.

Survey of experiments with ligated pylori on cat.

Cat

Exp.

Ale.
cone.
Per cent
by vol-
ume

Per cent alcohol
absorbed after

Diffusion

factor

Volume of
• contents of
stomach
ml

60'

120'

16.5

41.6

74-6

0.007 ± 0.002

Constant

0.017 + O.002

—

0.021 + 0.002

—

0.020 + O.OOl

—

0.023 + 0.002

—

0.018 ± 0.003

42.6

—

0.019 ± 0.002

—

0.019 ± 0.003

16.7

0.021 + 0.002

Diffusion factor vith ligated pylori (ale. 5 %) : average 0.018 ± 0.002 (8 exp.)
Alcohol absorbed after 30 min. :S8 !i (5 exp.)

> > >60 min. : 62 >

In the first series of experiments — with intact pylori — a
greater percentage of the alcohol had disappeared from the stom-
ach after a certain time than compared with the correspond-
ing values in the series with ligated pylori The difference will
probably be due to a partial emptying of the contents of the
stomach into the intestines — as indicated by the decreasing
values of the volume V — rather than to a more rapid absorption
in the cases of intact pylori, for in the experiments, where the
volume of the solution ingested had been constant throughout
the experiment, the magnitude of the diffusion factors, in aver-
age 0.017 ^ 0.003, was the same as in the cases with ligated
pylori: 0.018 ^ 0.002. Whether the alcohol after passage into
the intestines was absorbed more quickly or slowly cannot be
determined by this method.

The low values of the diffusion factors in the first experiments
in the two series (exp. 2 and 3) as compared with the average,
will most likely be due to inferior technique and also perhaps
to inferior circulation conditions in the stomach, which may
cause a lower diffusion factor.

ABSORPTION OF ETflTI. ALCOHOL.

263

B. Himinn Subjects.

L Normal Individuals.

As an extension of tlie results obtained on cat, a series of ex-
periments bave been carried out on human subjects in order to
see to what extent the above laws may be applied here.

The example below may serve to illustrate a normal experiment.
23. 5. 38. Exp. 10. H. J. 67 kg. 300 ml of an ale. sol. of
5 % by vol. by stom. tube.

Time

min.

AJc. cone.
®/o<l

«/oo

mm.

°/oo

Diff. fact.

32.G

29.6

3.0

31.1

0.019

28.2

1.4

28.9

O.OlO

25.8

2.4

27.0

O.0I8 Average : 0,016 + 0,002

24.0

1.8

24.9

0.016

21.9

2.1

23.0

0.018

19.2

2 7

20.6

0.026

17.1

2.1

18.2

0.028

15.0

2.1

16.1

Average : 0.026 ± 0.002

13.5

1.6

14.2

0.021

lO.G

2.9

12.1

0.024

7.9

2.7

9.S

0.032 1

Diffusion factor, before emptying : average O.OIG ± 0.002
j > after > > 0.025 ± 0.002

Alcobol absorbed after 30 min : 41 H
» > » 60 > :73

As -will be seen the diffusion factor shows good constancy up
to 30 min — 0.016 ± 0.002 — then to rise to 0.026 and after-
wards to assume a constant value but now on a somewhat higher
level: 0.025 ± 0 . 002 , the difference being statistically signific-
ant, Based on the experiments on animals with intact and
ligated pylori, it may be assumed that the reason for the sudden
change in the diffusion factor is a partial emptying of the stom-
ach with a subsequent decrease in the diffusion factor [see for-
mula (9)]. After this the volume seems to keep constant, for the
diffusion factor likewise remains constant.

2^ SVEN M. BERGGEEN AND LEONARD GOLDBERG.

An experiment vras also made on tlie same human subject
with a somewhat stronger alcoholic solution in the form of beer
with an addition of extra alcohol, corresponding to a solution
of 8 per cent by volume, This is illustrated by fig. 4.

Exp. 11. J. H. 67 kg. 330 ml beer + 15 ml ale. abs., corresponding to a solu-
tion of 8 per cent by volume, by stomach tube 8h.

0 — 0 — 0 alcohol percentage in contents of stomach
• — • — • logarithms for alcohol percentage
4" emptying of stomach at 90 min.

Diffnsion constant before emptying; 0.021 i 0.002
t » after a : 0.032 ± 0.008

ABSORPTION OF BTHYL ALCOHOL. 265

This experiment shows the agreement between the results at
repeated experiments even on human material. The diffusion
factor remains constant: 0.021 ± 0.002, for as long as 90 min,
then suddenly to rise to 0.034 and afterwards to remain more
or less constant with an average of 0.032 dr 0.008. As in exp.
10 the reason for this sudden change must be ascribed to a par-
tial emptying of the stomach with a corresponding decrease in
the volume of the solution ingested — about 33 % when 'cal-
culated from the change in the diffusion factor [formula (9)];
afterwards a constancy of the volume again appears, as indi-
cated by the constancy of the diffusion factor. — The sudden
change in the course of absorption which takes place after 90
min, is not clearly visible when forming a curve of the absolute
values representing the alcohol percentage of the contents of
the stomach, but only after calculating the diffusion factors (as
in exp. 10) or by forming the theoretical exponential curve by
means of the method of the least squares (curve o — o in fig. 4)
or by plotting the logarithmic values of the alcohol percentage
(•-•-• in fig. 4).

The following example where the diffusion factor increases
gradually, would then be connected with a continuous emptying
of the stomach.

3. 6. 38. Exp. 16. S. M. B. 80 kg. 300 ml of an ale. sol. of
® % t>y vol. through stomach tube S’*.

Time

min.

Ale. cone,
"/m

0/00

min>

o/oo

Diffusion factor

31.7

27.2

4.6

29.6

0.030

13.6

18.6

20.4

0.061

20 . . . ....

7.75

5.86

10.7

0.186

Alcohol disappearing after 30 min : 100

A survey of the normal cases is given in table 3. In certain
cases the values of the diffusion factors during the experiment
have shown a sudden increase. In analogy to what has already
been said this increase will be due to a partial emptying of the

266

SVEN M. BEROaKEN AND LEONARD GOLDBERG.

stomacli, and tlie time for this, calculated from the change in
the diffusion factor, is also to be seen from table 3,

Tal)Ie 8.

Survey of experiments with alcohol test meal on normal
individuals.

Subj.

Ale. cone.
Per cent
by volyme

Per cent ale. dis-
appearing from stom-
ach after

Stomach

emptied

after

Diffusion factor

30'

60'

120'

before

emp

after

tying

Differ-

ence

n— I

10'

0.029

0.090

0.061

25'

0.037

O.OfiO

0.023

0.083

25'

0.016

0.026

•0.009

90'

0.021

0.032

O.oil

100

0.080

0.098

0.068

39'

0.019

O.049

0.080

Diffusion factor before emptying : average 0.026 A: O.008 (6 exp.)

> > after » » 0.0B9 + O.OIS >

Alcohol disappearing after 30 min : 52 (5 exp.)

» » » 60 min :81 f6 j

From table 3 it is evident that the diffusion factors in the dif-
ferent individuals previous to the supposed emptying of the stomach
are of approximately the same magnitudes vdth an average of
0.025 0.003; after emptying, however, the values rise con-

siderably: in average to 0.059 i 0.013, The average of the in-
crease of the diffusion factors after emptying, when, calculated
from 6 experiments, is 0.034 ± 0.010, which is statistically sig-
nificant.

H. Patients.

Some experiments have been made on patients, especially those
suffering from achlorhydria and gastritis, with an attempt to apply
the above experiences on clinical material. The alcohol was

ABSORPTION OF ETHYL ALCOHOL. 2f>7

given as a test meal: 300 ml of an alcoholic solution of 5 per cent
by volume through stomach tube, and samples were withdrawm
everj' 5 — 10 min. The cases were kindly placed at our disposal
by doc. G. Xy'LIX, Serafimerlas., med. din. II. A survev of the
cases is given in table 4.

Table 4.

Sxirveit of expermenis iriih alcohol lest meal on patients.

Per cent ale. dis-

j ■'

■ appearing from

' Diffusion factor

T,., Ale. cone., stomach after

“"P- i Per cent !-

l»v volvme

.30' i 60’

Stomach

Snbj.

emptied 1

after ' before .after, piffer-
emptying once

-j ir-r

Lg 1

72 i

1 96 1

40'

0.03G

0.077

0.041

Lq ;

o f'

92 1

m and 40'j

0.030 '

0.08T

0.057

^ 1

6 I

—

0.050

s i

® 1

25’ i

0.03G

Ja 1

13 j

i 87 1

—

25' .

0.043

0.082

0.03!i

14 1

i 46 i

47' ,

0.018

0.O75

0.057

Diffasion factor before

emptying

; : average

0.08C ± O.OOG (4 exp.)

after

0.080 + 0.

003

Alcohol disappearing after 30 min : 62 (4 exp.)

> » >60 min: 92 X >

According to table 4, there seems to be no essential devia-
tion from the results found on normal individuals. The figures
show a certain tendency to higher diffusion factors among the
clinical material, which may possibly be due to a more rapid
disappearance of the alcohol from the stomach. There may be
two reasons for this, firstly a more rapid absorption, something
not altogether unusual in the case of gastritis, secondly a more
rapid disappearance owing to the stomach emptjnng more ra-
pidly, this being a well known fact in the case of achlorhydria.
The material is how^ever. insufficient to allow of statistics, and
should be completed with absolute figures of the volume of the
contents of the stomach if it is to be a guide as to w’^hich of the
possibilities predominates.

19 — i01323. Acta phys. Scandinav. Vol. 1.

268

SVEN M. BERQQREN AND LEONARD GOLDBERG.

Conclusions.

By continuously following the variations in the alcohol per-
centage of an alcoholic solution ingested in the stomach, it has
been possible to obtain the curve for the disappearance of alco-
hol from the stomach. In cases where the pylorus has been li-
gated or repeated measurements have proved convincing that
no change has taken place in the volume of the solution ingested,
the disappearance of alcohol must be ascribed to absorption.
Assuming that absorption follows the laws of diffusion and taking
into consideration the volume and concentration of the solution
ingested, the percentage of blood alcohol and the blood flow
through the mucous membrane of the stomach, it has been pos-
sible to obtain a so-called dijjusion factor which, provided the
volume ingested and the blood flow are constant, will likewise
be a constant. Although there are some variations in the size
of the diffusion factor, a survey of all the cases shows its constan-
cy and thereby the tenability of the assumption, to which the
experiments have been adapted, i. e. that the passage of ethyl
alcohol from the gastro-intestinal tract to the blood is a simple
process of diffusion. The deviations from the constancy of the
diffusion factor can equally be explained by the changes of the
other elements' forming this factor. The formulae set up have
also proved ajiplicable on human subjects.

Summary.

After a survey of earlier literature on the passage of ethyl
alcohol from the gastro-intestinal tract to the blood, the factors
are discussed that may possibly have to bear on this passage,
the basis of the discussion being the laws of diffusion. Prom
these laws formulae have been deduced, in which the percentage
of the blood alcohol, the blood flow through the mucous mem-
brane of the stomach and the concentration and the volume of
the alcoholic solution ingested have been taken into considera-
tion. The validity of these formulae is not only confined to the
diffusion of alcohol but also applies to all the cases in which
a pure process of diffusion is concerned.

By means of a series of determinations of the alcohol percent-
age in the contents of the stomach of cat with intact and ligated

ABSORPTION OF ETHYL ALCOHOL.

2()9

pylori as -well as on luiman subjects — botli normal individuals
and patients — by means of fractional Avithdrawal of ingested
alcohol, an attempt has been made to show that the passage
of alcohol through the mucous membrane of the stomach is to
be regarded as a pure process of diffusion. The results of the
ex])eriments arc;

1) The absorption of ethyl alcohol from the stomach follovs
the laws of diffusion. This means that the greater the con-
centration of the alcohol ingested, the more rapidly will
absorption take place with a higher and earlier blood alcohol

a (1 — k)

maximum (cp fig. 1). The so-called diffusion factor

is constant in the case of constant blood flow and constant vol-
ume.

2) A cat with ligated ])ylorus absorbs bi 30 min about 38 %
of the alcohol ingested, when given in a solution of 5 per cent
by volume, about G2 % after 60 min and 76 % after 120 min;
of a solution of 17 per cent proportionally more is absorbed.

3) On human subjects with closed pylori the corresponding
figures for a solution of 5 per cent by volume are 40 % after
30 min, 70 % after 60 min and about 100 % after 120 min.
If the stomach is emptied, on the other hand, the whole quantity
of alcohol ingested can disappear in less than 30 min.

References.

Abramson, L., and P. . Linde, Arch. int. Pharinacod}Ti. 1930. S9.
32.5.

Berggren, S. M., To be published.

Bernhard, C. G., and L. Goldberg, Acta med. scand. 1935. oo. 15...

Braxdl, J., Z. Biol. 1893. 11. Til. • j r i i

Le Breton, E., Signification physiologique de boxydation de 1 alcool
ethylique dans Torganisme, Lons-Le-Saunier 1936.

Carpenter, T. M., J. Pharmacol. 1929. 37. 217.

— , Scientific monthly 1937. 45. 5.

CORI, c., E. L. ViLLiAUME and G. Cori, J. Biol. Chem. 1930. S7. 19.
Edkins, N., and M. Murray, J. Physiol. 1926. 62. 13.

Elbel, H., and G. Lieck, Z. ges. ger. Med. 1936. 26. 270.

Faure, W., and S. Loewe, Biochem. Z. 1923. 143. 47.

Fleming, R., and E. Stotz, Arch. Neurol. Psychiat. Chicago 1935.

33 ^02

Gettler, A. O., and A. W. Freireich, Amer. J. Surg. 1935. 27. 328.
Graf, 0., and E. Flake, Arbeitsiihysiologie 1933. 6. 141.

270

SVEN M. BER6GREN AND LEONARD GOLDBERG.

Grehant, N., C. R. Acad. Sci, Paris 1895. 120. 1154.

Haggard, H. W., and L. A. Greenberg, J. Pharmacol. 1934. 52.
167.

Handwerk, W., Pharm. Beitr. AUcoholfr. 1927. H. 2.

Haneborg, a. 0., The effects of alcohol upon digestion in the stom-
ach, Christiania 1921.

Hanzlik, P. J., and R. J. Collins, J. Pharmacol. 1913. 5. 185.
Harger, R. N., H. R. Hulpieu and E. B. Lamb, J. Biol. Chem. 1937.
120. 689.

JuNGMiCHEL, G., Alkoholbestimmung im Blut, Berlin 1933.

— , Arch. exp. Path. Pharmak. 1933. 173. 388.

Kautsch, M., tlber die Resorption einiger Alkohole von verschiedenem
Siedepunkt im Magen, Halle 1898.

Linde, P., Arch. exp. Path. Pharmak. 1932. 167. 285.

Lucas, A., Arch. Verdau Rr. 1930. 48. 332.

Lonnquist, B., Bidrag till kannedomen om magsaftutsondringen,
Helsin^ors 1906.

Matossi, R., Z. klin. Med. 1931. 119. 268.

Mellanby, E., Medical Res. Comm. Spec. rep. 1919. 31.

Miles, W. R., J. Pharmacol. 1922. 20. 265.

Nemser, M. H., Hoppe-Seyl. Z. 1907. 53. 356.

Newman, H. W., and J. Card, J. Nerv. Ment. Dis. N. Y. 1937. 86.
428.

Neymark, M., and E. M. P. Widmark, Skand. Arch. Physiol. 1936.
73. 260.

Nicloux, M. C. R. Soc. Biol. Paris 1931. 107. 997.

— Bull. Soc. Chim. Biol. Paris 1934. 16. 330.

Nyman, E., and A. Palmlov, Skand. Arch. Physiol. 1936. 74. 155.
Glow, J., Biochem. Z. 1923. 134. 553.

Schmidt, M., J. industr. Hyg. 1934. 16. 355.

— , Alkoholsemi, Kobenhavn 1937.

ScHWEiSHEiMER, W., Dtsch. Arch. klin. Med. 1913. 109. 271.

Segall, M., Versuche iiber die Resorption des Zuckers im Magen,
Miinchen 1888.

Southgate, H. W., Biochem. J. 1925. 19. 737.

Tappeiner, H. von, Z. Biol. 1880, 16. 497.

Tuovinen, P. I., Skand. Arch. Physiol. 1930. 60. 1.

Widmark, E. M. P., Ihidem. 1916. 33. 85.

— Rungl. Fysiogr. Sallsk. handl. 1930. N. P. 41. no. 9.

— , Die theoretischen Gnmdlagen und die praktische Verwendbhrkeit
der gerichtlich-medizinischen Alkoholbestimmung. Berlin und
Wien 1932.

— , Biochem. Z. 1934. 270. 297.

VoLTZ, W. U. und W. Dietrich, Biochem. Z., 1915. 68. 118.

From the Physiologj’ Laboratory, Helsingfors University
and the Neurophysiological Department, The
Caroline Institute, Stockholm.

Evidence for two Phases in the Regeneration
of Yisual Purple.^

M. ZEWI.

(With 5 figures in the text.)

It was noted by Zewi (1939) that, although the process of
regeneration of ^dsual purple in live frogs had a high temperatiwe
coefficient, this process in excised opened eyes could not be
accelerated by a rise in temperature. Kegeneration in isolated
eyes was also relatively slow, and, in order to obtam an equa y
slow rate of regeneration in the live animal, the temperature a
to be brought dovm to about 8°. The isolated eyes thus behaved
as if they had been lacking that phase of regeneration w
the intact frog is particularly sensitive to temperature.

Some observations on the effect of pilocarpine an a ropme on
V. P. (visual purple) regeneration are of interest in this connemon
and will therefore be reported. They were made m order to ^
with a quantitative technique some old ^ .

and Ktjehxe (1882) on rabbits and by Dbeseb (1886) on g .

Pilocarpine was used by them on account of its ® ,

on secretion and atropine as the corresponding in i i
stance. This type of problem and the general ^ f ®

dravm from experiments of this kind do not appear o u
same light to-day as in 1880. Nevertheless the
justify a repetition of the experiments with t e n+Tonine

nique now at our disposal. The old authors
had no effect on the regeneration of visual purp e u
carpine enhanced it. These observations were confirmed by

Amenomiya (1930).

* Received 17 September 1910.

272

M. ZEWI.

Fig. 1. Visual purple regeneration in live frogs at 22.4°: • after injection of 5 mg
])ilocarpine (2.3 eyes), r after injection of 5 mg atropine (28 eyes). C controls
(22 eyes). Abscissae: time in tlie dark in hrs. Ordinates: density of visual purple.
Later figg. similarly marked.

Technique.

As experimental animals served Hungarian frogs. The visual
purple Tvas extracted vdth 2 per cent digitonin (Taxsley, 1931).
Each retina was separately extracted with 1 ccm of the solvent
and left in it for 40 min. at room temperature. Then followed
centrifugation for 20 min. at 3,800 rev./min. Of the clear extract
0.4 ccm was measured into an absorption trough of 20 mm length.
This was put into the beam of a light of wave-length 0.498 /t
and its absorption determined with the aid of a photocell and an
electrometer. For details, see Graxit, Holmberg and Zewi (1938)
and Zevt (1939).

KEGENERATION OF VISUAL PURPLE.

278

Fig. 2. Visual purple regeneration in live frogs at
pilocarpine (31 eves), r after injection of 5 mg

(21 eyes).

8.0': • after injection of 5 mg
atropine (29 eyes). C controls

Visual purple concentration will be expressed in tlic usual man
ner as the difference in density ^which is log transmitted/ ^
the original unbleached extract and the fully bleached extract.

Results.

1. Regeneration in live frogs.

Frogs dark adapted overnight were light adapted for one hour
to 20,000 m. c. Immediately after light adaptation some anima s

received 5 mg. pilocarpine into the lymph sac, ®

amount of atropine, and some served as controls. The ogs vere
then left to dark adapt in a water bath at constant
(Zewi, 1939). They Avere removed from the water ba
suitable intervals in the dark, decapitated in re ig ’
which the eyes were analyzed for visual purp e. n is m
regeneration curves were obtained for two tempera ure ,
and 8° C.

274

M. ZEWI.

Such curves are shown ih figg. 1 and 2. They are not drawn
from origo but from a point corresponding to a V. P. concentration
of 0.040, which is the average minimum to which visual purple
can be brought by light adaptation (Zewi, 1939). The readings
marked in the figures are averages of several eyes.

Pig. 1 shows that regeneration in all cases initially foUow's a
common course but that later on pilocarpine has slightly accel-
erated the rate of formation of visual purple. Atropine very def-
initely slows down regeneration.

Pig. 2 gives the additional significant information that the
slew rate of regeneration at 8° is uninfluenced by either drug.

2. Regeneration in opened isolated eyes of animals injected
with pilocarpine or atropine.

The dark adapted frogs received an injection of 5 mg. pilocar-
pine or atropine and were then light adapted for one hour to
20,000 m. c. After this the eyes were removed, opened, and
left to dark adapt at 22.2°.

Pig. 3 shows that pilocarpine had no effect whatsoever, but
that, contrary to expectation, atropine now very definitely
improved the rate of V. P. regeneration.

3. Regeneration in opened isolated eyes treated
with pilocarpine or atropine.

The eyes of dark adapted frogs were removed and opened in
red light and thereafter illuminated for 30 min. with a 1,000 W.
lamp at a distance of 40 cm. After light adaptation in this manner
the right eye received one drop of a 1 per cent solution of atropine
or pilocarpine. The left eye served as control and had a drop of
Ringer. Then followed dark adaptation at 22.4° and the amount
of visual purple was determined for each pair of eyes in parallel.
This procedure is based on the fact that V. P. concentration
in right and left eye does not as a rule differ by more than, on an
average, 1 per cent, provided that a sufficient number of eyes be
used (Geanit et al. 1938, Zewi 1939).

Fig. 4 shows that regeneration was suppressed after pilocarpine.
The less definite result of fig. 5 illustrates that atropine suppresses
regeneration.

REGENERATION OP VISUAL PURPLE.

275

Before

atropine

lymph sac. Controls are maraea j.
Vere used for the respective senes.

Fig. 4. Visual purple regeneration in 18 rSed^only I drop

with pilocarpine • compared with_ 18 controls o, ^

of Ringer.

276

M. ZEWI.

Pig. 5. Visual purpjle regeneration in IS excised and opened eyes at 22.4° treated
with atropine •, compared with 18 controls o, which received only a drop

of Einger.

Discussion.

The interesting results are those of figg. 1 and 2 in which a
differentiation of the regenerative component processes was carried
out by means of temperature. At 8° neither drug had any in-
fluence but at 22. 4° pilocarpine somewhat accelerated and atropine
definitely slowed down the rate of regeneration of visual purple.
This must be regarded as further evidence in favour of the dual
nature of Y. P. regeneration. Only the component sensitive to
temperature is influenced by the drugs. It is hardly probable,
despite the low temperature used, that within the 5 hours during
which the process was followed, absorption of the drugs from the
lymph sac should not have taken place.

Very curious is the fact, shown in fig. 3, that excision of the eye
of the atropinized frog leads to an improvement in the rate of
regeneration of visual purple as compared with control eyes.
The same improvement did not appear in the live frog at 8°. It
reminds one of observations by Jarvi (1935) according to which the
secretory cells of trachea and larynx of cats and guinea pigs after
atropinization are richly filled with secretory substance although
secretion outwards cannot take place. Atropine may favour some
preparatory process, say, in the retinal pigment cells but this
favourable effect cannot lead to increased formation of Nusual

REfiBNERATION OF VISUAL POSPLE. -fll

purple before tlio eye is removed aud thereby withdrawn from the
direct influence of the drug.

Very little use can be made of the circumstance that both
drugs impede regeneration of visual purple when dropped into
excised opened ej'es. This maj' interfere directly with the state
of the tissues which for good regeneration must be normal (Zewi,
1939).

I am indebted to Professor Eagkar GRaxit, Stockholm, for
technical support and criticism of this work as well as for its
translation into English.

Summary.

The concentration of visual purple has been determined photo-
electrically during dark adaptation of previously light adapted
frogs at 22.4 and 8°.

At the higher temperature pilocarpine slightly enhances, and
atropine definitcl}’ suppresses the rate of regeneration of visual
purple in live frogs. At the lower temperature neither substance
has any effect on the regenerative processes.

The results are interpreted as further e\’idence in favour of
the dual nature of visual purple regeneration. Only the process
sensitive to temperature is in this particular instance sensitive
to the drugs.

Some results with regeneration in excised eyes under the
influence of pilocarpine and atropine are also reported.

References.

AjfEN'OMiVA, J., Actn Soc, Ophthal. jnp. 1930, 34, IISO; quoted from
Ber. gcs. Physiol. 1931, 60, 124.

Ayres, tV. C., and W. Kuehne, Unters. physiol. Inst. Heidelberg 1882,
2, 215.

Boll, P., Honatsber. Akad. Wiss. Berlin 1876, p. 783.

DrESER, H., Z. Biol. 1886, 22, 23. n, Aon

Grakit, R., T. Holmberg and M. Zewi, J. Physiol. 1938, 94, 430.
Jarvi, 0., Annal. Acad. Scient. Penn. A. 1935, 43.

T.\xsley, Iv., j. Physiol. 1931, 71, 442.

Zewi, HI., Acta Soc. Scient. Penn. X. S. 1939, B. II, 4.

Aus der Medizinisch-Chemisclien Abteilung der TJniversifat zu Lund
und der medizinischen Abteilung des Kreiskrankenhauses Orebro.

Eine eiiifaclie und iur klinische Zwecke geeig-

nete Mikrometliode ziir Bestinimimg des Harn-

stoffstlckstoffs (UrN) im Blute durcli Urease-
beliandlttug imd direktc ^^esslerisierimg.^

Von

WILHELM OHLSSON,

Kay und Eeid (1934) haben nacbgewiesen, dass der optimale
pH-Wert flir die Urease zwischen 6.6 und 7.0 liegt, und aus ibren
Versucben geht hervor, dass noch bei einem pjj von 6.2 fast op-
timale Enzyravirkung erzielt ’wird. Bei den bisherigen Metboden
zur Bestimmung des Harnstoffs mit Hilfe von Urease bat es sicb
im allgemeinen denn aucb als notwendig erwiesen, eine besondere
Pufferlosung zu verwenden, um einen geeigneten pH-Wert zu
erbalten.

Bei der von mir scbon friiber bescbriebenen Metbode zur Be-
stimmung des Keststickstoffs (1937) ^vird die entnommene Blut-
probe in eine sog. Molybdatlosung gebracht, die Natriummolybdat
und Kaliumsulfat entbalt. Diese Molybdatlosung, deren eigent-
liche Aufgabe es ist, als eiweissfallende Losung zu dienen, hat ein
Ph von 6.2 — 6.5 und erwies sicb als ein ausgezeicbnetes Medium
flir die Ureasewirkung. Die aufgefangene Blutprobe kan desbalb
direkt mit Urease versetzt werden, und das besondere Zusetzen
einer Pufferlosung eriibrigt sicb.

Das Prinzip der Metbode besteht darin, dass der Harnstoff des
Blutes mit Eblfe der Urease in Ammoniumkarbonat ubergefiibrt
wrd, dessen Stickstoffmenge, nacb Eallen des Eiweisses, mit dem
Kessler ’scben Reagcns direkt kolorimetriscb bestimmt werden
kann.

’ Der Redaktlon am 28. September 1940 zngegangen.

BESTIMMUKG DBS IIARNSTOFFSTICKSTOFFS. 279

Die iMelirzalil dcr fiir diese jMethode erforderliclien Losungen
stimmt rait den bci der oben erwabnten Jletbode zur Eeststickstoff-
bcstimmung zur Ver^venduug kommenden iiberein. Es muss nur
noch cine Urease-Losung und eine Gummi-arabicum-Losung
bcreitet werden.

Erfordcrliclie liosungen.

Mohjbdatlosung: 10 ml 10 %ige Katriummolybdatlbsung (pro
analysi) uud 6.5 g Kaliumsulfat (pro analysi) werden in einen
500 ml fassenden l\Iesskolben gebrncht und bis zur Marke mit
destilliertem Wnsser verdiinnt.

FaUtnigslosung: 40 ml 1-normaler Scb^Yefelsaure -ft-erden in
einen 300 ml fassenden Messkolben gebracht und bis zur Marke
mit destilliertem Wasser vordiiimt.

Siandardlosungcn: 0.4716 g getrocknetes Ammoniurasulfat
(pro analjsi) werden im Messkolben in destilliertem Wasser bis
auf die itengc von 1 Liter gelost. Die Losung enthalt 0.1 mg
Stickstoff pro ml.

Standard 1: 20 ml der obigen Standardstammlosung -werden
auf 200 ml verdiinnt. Enthalt 0.01 mg Stickstoff pro ml.

S(a7]dard II: 80 ml der obigen Standardstammlosung werden
auf 200 ml verdiinnt. Enthalt 0.04 mg Stickstoff pro ml.

Stamrnldsung fiir das Ncsskr’sche Reagetis: In einem J enakolben,
der 500 ml Wasser von 70° enthalt, werden 90 g Jodkalium und
100 g Quecksilberjodid gelost. Die Losung wird abgekiihlt, in
einen 1 Liter fassenden ^klesskolben gebracht und bis zur Marke
verdiinnt.

Nessler’sclics Reagens: 450 ml der oben erwahnten Nessler’schen
Stammlosung werden mit 2 100 ml 10 %iger Hatronlauge imd
1 200 ml destilliertem Wasser versetzt, worauf sorgfaltig gemischt
-wird. Die Losung wird in dunkler Flasche aufbewahrt.

Samtliche oben beschriebenen Losungen sind dieselben, die
bei der oben erwahnten Methode zur Reststickstoffbestimmung
zur Yerwendung kommen.

Ureaseldsung : 1 Tablette Urease Squibb (E. R. Squibb a. Son,
New York) zu 0.1 g wird zerdriickt und in 1 ml destilliertem Wasser
aufgeschwemmt. Diese Losung wird bei Vornahme der Bestim-
mung jedesmal frisch bereitet.

Gummi-arabicim-Losung: 1 %ig.

280

■\VILHELM OHLSSON.

Ausfuhrung.

Behandlung der Blutprohe: In ein Zentrifugenglas, das 8 ml
Molybdatlosung enthalt, werden 0.2 ml Blut gebraclit, wobei die
Blutpipette durchgespiilt wird. Es werden 2 Tropfen Ureaselbsung
hinzugegeben. Sorgfaltiges Mischen durch vorsichtiges Schiitteln.
Stehenlassen der Mschung 30 IVIinuten lang bei Zimmertempera-
tur. Darauf werden 2 ml der Eallungslosung binzugefiigt und nach
Mischen, 10 Minuten lang Icraftig zentrifugiert. Das klare Zentri-
fugat wird dekantiert, z. B. in ein anderes Zentrifugenglas, und
8 ml da von werden in ein Reagenzglas gebracbt, worauf 2 ml
Gummi-arabicum-Losung binzugefiigt werden.

Standardprobe: 4 ml Standard I und 4 ml Standard II werden
in je ein Reagenzglas geschiittet. Zu den Proben werden je 2 ml
Eallungslosung und 4 ml destilliertes Wasser binzugefiigt. Sorg-
faltiges Miscben.

Sowobl die Blutprohe als aucb die Standardprobe werden mit
je 5 ml Nessler’scbem Reagens versetzt, worauf man im Kolori-
meter vergleicbt. Kommt die Standard-I-Probe zur Verwendung,
und wird sie auf 20 eingestellt, so findet die Berecbnung in fol-
gender- Weise statt:

10 X 1 000 X 20 X 0.04
8 X 2 xP

B = mg UrN pro

100 ml Blut;

500

— B

= mg UrN pro 100 ml Blut;

P = An der Probe am Kolorimeter abgelesener Wert.
B = Blindwert in mg % N.

Vor allem auf Grund der Gegenwart von Gummi-arabicum-
Losung muss jedesmal, wenn diese Losung frisch bereitet worden
ist, die Blindwertbestimmung vorgenommen werden. Dieser
Blindwert, der im allgemeinen nicbt fiber 1 — 2 mg % betragt,
wd in folgender Weise bestimmt; In ein Reagenzglas, das 4 ml
Standard I entbalt, werden 2 ml Gummi-arabiQum-Losung,
2 ml EMlungslosung und 2 ml destilliertes Wasser geschiittet.
Es wird einc Standard I entbaltende Standardprobe bereitet
(siebe oben!). Beiden Reagenzglasern werden je 5 ml Nessler’sches
Reagens zugesetzt, worauf gemischt wird. Die Standard-I-Probe

BESTIMMU^'G DBS IIAIIKSTOFFSTICKSTOFFS.

281

wird auf 20 mm eingestcllt, und daim die Bliiidprobe dagegen
kolorimetriert, wobei der Blindwerfc in folgender Weise erhaltcn
wird;

1 000 1"^ = B in mg % X;

P, = an der Blindprobe am Kolorimeter abgelesener "Wert.

Bei Werten des Harnstoffstickstoffs von uber 80 mg % findet
die Ablesung gegen die Standard-II-Probe statfc, und die Berech-
ming \vird dann folgende sein:

10 X 1 000 X 20 X O.IG
8 X 2 X P

— B = mg UrN pro

100 ml Blut;

2 000

p- =

IJrN pro 100 ml Blut;

Bei der Bereclmung wcrdcn die zugefiihrte Blutmenge (0.2ml)
und die Ureasemenge (2 Tropfen) nicht bcriicksichtigt. Das Ver-
nachlassigen dicser Paktoren fuhrt zu einem Felder von etwa 2 %,
was sich jedoch im cndgiiltigen "Werte kaum bemerkbar macht.

Die Ureasemenge wurde mdglichst klein gcwahlt (2 Tropfen
einer 10 %igen Losung), da es sich herausgcstellt hat, dnss selbst
frisch bereitete Ureaselosungen nicht ganz unbedeutende Mengen
NHj enthalten. Diese geringe Ureasemenge hat sich jcdoch selbst
fiir die grbssten im Blute denkbaren Harnstoffmengen als geniigend
erwiesen, und ihr eigener NUj-Gehalt kann ganzlich vernach-
lassigt werden, falls die Losung jedesmal, wenn eine Bestimraung
vorgenommen werden soil, frisch bereitet wird. Selbst wenn die
Losung im Eisschrank aufgehoben wdrd, steigt ihr NHs-Gehalt
ziemlich rasch.

Wiihrend man bei der Methode zur Reststickstoffbestimmung

mit 0.1 ml Blut arbeiten kann, hat es sich bier als notwendig

erwiesen, 0.2 ml Blut zu verwenden, um bei den normal vorkom-

menden etwa 15 mg % UrN mit dem Nessler’schen Reagens eine
fiir die Kolorimeterbestimmung geeignete Farbenstarke zu er-
halten.

Als Schutzkolloid hat sich Gumim arabicum als uberaus wirk-
sram erwiesen. Seine Gegenwart bringt es jedoch mit sich, dass,
A\ie oben erwahnt, eine Blindwertbestimmung vorgenommen "wer-
den muss.

WILHELM OHLSSON.

Was die Normahverte des Harnstoffstickstoffs im Blute anbe-
langt, sind diese, nacb. Angabe mehrerer Autoren, nicht nur indi-
viduell verscbieden, sondern. scbwanken auck beim gleicken Men-
schen innerhalb gewisser Grenzen. So geben. z. B. Cameron und
Gilmour (3) 8 — 20 mg % an, Thumper und Cantarow (4)
12 — 15 mg %. MacKay und MacKay (5) finden bei Frauen
niedrigere AVerte (5.14 — 18.2 mg %) als bei Mannern (12.05 —
21.6 mg %). Ein Normalwert von 9 — 15 mg % diirfte jedoch der
am haufigsten angenommene sein.

Normalerweise macht der Harnstoffstickstoff 50 — 60 % des
Reststickstoffwertes aus. Bei Niefeninsuffizienz wird jedock ver-
kaltnismassig mekr Harnstoffstickstoff als andere Reststickstoff-
bestandteile zuriickgekalten, wodurck seine Menge auf iiber 60 %
steigt. Besonders bei leichteren Graden von Niereninsuffizienz
diirften deshalb Parallelbestimmungen des Reststickstoffs und
des Harnstoffstickstoffs von gewissem AVert sein. AVenn man,
Tvie hier, bei beiden Bestimmungsmetkoden im grossen Ganzen
mit den gleicken Losungen arbeitet, bringt dieses den Vorteil
mit sick, dass die AA^erte untereinander besser vergleickbar sein
werden.

Tabelle L

Mit den hier angegebenen Meihoden ausgefiihrte Parallelbestimmungen
des Reststickstoffs und des Harnstoffstickstoffs.

1 RX mg %

UrK mg %

UrN X 100
RN ■

53.8

’ 9

14.6

50.0

56.2

45.5

73.3

42.4

42.8

50.0

151

116

76.8

i 9

40.G

Um von der Zuverlassigkeit der Metkode ein Bild zu bekommen,
stellte man folgenden Kontrollversuck an: Je 0.1 ml Serum mit
bekanntem Gekalt an Harnstoffstickstoff (14.1 mg %) wurde in
Zentrifugenglaser gebracht, die je 4 ml Molybdatlosung enthielten,

BESTUIMUNG DBS HARNSTOFFSTICKSTOFFS. 283

in der verscliiedeii grosse, bekannte Mengen Harnstoff gelost
■waren. Die Probeii wurden in der Polge in volliger tjbereinstim-
mung mit dem oben beschriebenen Verfabren bebandelt, nur mit
dem Unterschiede, dass bier nur die balbe Menge Molybdatlosung
und Pallungslosung zur Verwendung kam.

Tabcllc n.

Berechncte
Menge hinzn-
gegebenen

UrX, mg %

Gefnndene

Menge

UrN, mg %

Gefandene Menge
+

UrN, mg % abziig-
licli der im Sernm
vorhandenen 14.1
+

mg % IJrN

Abweicbnng in

mg %

22.1

35.9

21.8

O.G

2.7

22.4

36.3

oo o

0.2

0.9

OO A

36,1

22.0

0.4

1.8

OO A

36.3

22.2

0.2

0.9

57.4

71.1

57.0

0.4

0.8

57.4

70.4

56.3

1.1

2.3

57.4

70.1

56.0

1.4

3.0

57.4

71.4

57.3

0.1

0.2

S0.8

94.0

79.9

0.9

l.S

S0.8

94.9

80.8

94.3

80.2

0.6

0.85

80.8

94.0

79.9

0.9

1.3

150.8

168.7

154.C

3.8

2.7

150.8

166.3

152.2

1.4

1.0

150.8

165.1

151.0

0.2

0.14

150.8

165.1

151.0

0.2

0.14 1

290.8

300.6

286.5

4.3

1.6 !

290.8

300.0

286.5

4.3

1.6

290.8

304.8

290.7

0.1

O.os

290.8

307.0

292.9

2.1

0.75 1

Zusammcnfossnng.

Es wird eine einfacbe Methode zur Bestimmung des Harnstoff-
stickstoffs an 0.2 ml Blut besclirieben. Ihr Vorzug besteht darin,
dass das Blut in eine Losung gebracbt wird, die gleicbzeitig als
eine fiir optimale Ureasewirkung geeignete Pufferlosung und nacb
leicbter Ansauerung als Ehveissfallungsmittel dient. Man erbalt
20 — i01323. Acta phj/s. Scandinav. Vol. I.

284

WILHELM OHLSSON.

ein eiweissfreies Zentrifugat und nimint darauf, in Gegenwart von
Gnmmi arabicum als Sclintzkolloid, nacL. dixekter Nesslerisierung
eine kolorimetrisclie Bestimniung vor.

Fiir einen Geldbeitrag zux Ausfiihrung der Arbeit mochte ick
der »Kungl. Fysiografiska Sallskapet» in Lund meinen Dank aus-
sprecben.

Literatur.

Cameeon, a. F.j und C. R. Gilmotjr, The biocbemistry of medicine,
London 1935.

Kay, W. W., und M. A. H. Reid, Biocbem. J. 1934. 28 . 1798.
MacKay, E. M., und L. L. MacKLay, J. din. Invest. 1927. 4. 127.
Ohlsson, W., Skand. Arch. Pliysiol. 1937. 75 . 207.

Thumper, M., and A. Cantaeow, Biochemistry in internal medicine,
Philadelphia and London 1932.

Department of Medical Chemistry, University of Helsinki.

In Yiti’o Studies on tlie Role of Vitamin D
in the Metaholisni of Calcium and
Pliosphorus in the Rat Bones.^

Y. V. KRAEMER, B. LANDTMAN and P. E, SHIOLA.

The object of the present work, which forms a part of a more
extended research on the mineral metabolism of the animal
organism, was to elucidate, on the basis of in vitro experiments,
the influence of vitamin D on the calcium and phosphorus metab-
olism in the animal organism.

Although the literature concerning the relation of vitamin D to
the metabolism of calcium and phosphorus is very extensive, up
to the present only a few studies have been recorded which try
to explain, with in vitro experiment, the role of vitamin D in
the processes of ossification, a question which still lacks more
detailed information.

Among the commrmications found in literature must be men-
tioned the observations made with in vitro experiments by Shtp-
iiEY, Kkamer and Howland (1925, 1926) that ossification may
occur in the rachitic cartilage, if the surrounding incubation solu-
tion contains a suitable amount of calcium and phosphorus.
Eobison and Soames (1930) mention that rachitic cartilage cal-
cifies in a regular manner, if placed in solutions of inorganic salts
supersaturated with respect to the bone salt, as well as in lower
concentrations of calcium and phosphorus if organic phosphoric
ester is present. Accordingly, this would indicate that rachiris is
a disease of blood and not of the bones. In later studies in vitro
Eobison and Eosenheim (1934) established that addition of

* Received 4 October 1940.

286

V. T. KEABMER, B. LANDTMAN AND P. E. SIMOLA.

vitamin D into the incubation solution had no histologically
demonstrable action on the calcification of the normal bone. On
the other hand, however, Kosenheim (1934), mentions that in the
in vitro experiments the calcifying power of the cartilage of rats
decreases with increasing periods on a rachitogenic diet. There is
also a short communication by FLEiscHiiANN (1937) that, accord-
ing to a histological study, the calcification of the bird bone occurs
at a slower rate in the serum plasma of rachitic animal than in the
normal plasma and that an addition of calcium and phosphorus
salts into the rachitic plasma does not suffice to compensate fully
the deficiency of vitamin D.

The experiments in vitro that have been performed with a
technique of more chemical nature, have chiefly dealt with the
question of the resorption of calcium and phosphorus from the
intestine. Some time ago Nicobaysen (1937) concluded from his
in vitro experiments that vitamin D promotes the resorption of
calcium from the intestine, but not the resorption of phosphorus.
In other tissues no difference seemed to exist in the absorption
of calcium by normal and rachitic animals. Harris (1932) had
already earlier advanced a theory that rachitis involves a failure
in the “net absorption” of calcium and phosphorus.

The above experiments for the elucidation of the role of vitamin
D were made only with organs of animals suffering from a defi-
ciency of vitamin D. Literature does not report chemical in-
vestigations which woxdd give more information based on experi-
ments in vitro concerning the metabolism of calcium and phos-
phorus in D-hypervitaminotic animals.

With regard to the lack of experiments in vitro concerning D-
hypervitaminosis we considered it reasonable to take this very
point for the chief object of our investigations which were started
in 1936. Therefore we used in the first experiments animals in
which D-hypervitaminosis had been induced by large amounts of
vitamin D. Somewhat later we began to arrange, together “with
experiments on D-hypervitaminosis, also experiments with D-
avitaminotic animals. The main purpose of all these experiments
was to find out whether it is possible to demonstrate, with experi-
ments in vitro under suitable conditions, the effect of \dtamin D
upon the ability of bone, firstly, to take up, and secondly, to retain
calcium or phosphorus.

In the following are presented the results of some of the prelim-
inary experiments made exclusively with rat bones.

STUDIES ON THE ROLE OF VITAMIN D.

287

Methods.

The experimental material consisted generally of preparations
obtained from the fore and hind legs of rats. In the experiments on
D-hypervitaminosis adult animals were used, the weight of which
varied from 150 to 210 g. The fore leg preparations (humerus,
radius and ulna together) weighed from 450 to 700 mg, the hind
leg preparations (femur and tibia) from 800 to 1 600 mg. The rats
in the experiments on I) ardtaminosis weighed from 40 to 80 g,
their fore leg preparations from 300 to 800 ing and the hind leg
preparations from 700 to 1 500 mg.

D-hypervitaminosis was produced by administration of varying
amounts of “vigantol” to the rats by means of a stomach tube,
the daily dosage varying from 12 000 to 36 000 I. U. The control
animals received the corresponding amount of vitamin D-free
sesame oil. The basal diet was otherwise the same in both
groups.

In the experiments on D-hypovitaminosis the diets of Mac
CoLLUM No. 3143 and Steenbock No. 2965 were used for the
production of rachitis. The control animals were kept free from
rachitis by giving them daily small doses of strongly diluted
“vigantol”.

The bone preparations were made as follows. Immediately after
killing of the animal the leg was carefully removed from the
surrounding tissues, care being taken to prevent damage of car-
tilage and periosteum. Tlie bone preparation was then placed in a
20 cc. test tube, which contained 10 cc. calcium cloride, sodium
phosphate or physiological sodium chloride solution respectively.
In experiments on the uptake of calcium we used in the in-
cubation calcium chloride solution containing 10—12 mg %
calcium and respectively, for the phosphor investigations sodium
phosphate solution containing 10 — 100 mg % phosphorus. In
both cases a suitable addition of sodium chloride was made into
the incubation solution to bring about isotonicity. In examining
the excretion of calcium and phosphorus from the bone 0.85 %
sodium chloride solution was employed as the incubation medium.

The bone preparation was incubated at room temperature for
24 hours.

In the analysis of the incubation solution the well-known
titrimetric micromethod of Kramer-Tisdall proved the most

288 V. V. KKABMER, B. LANDTMAN AND P. E. SIMOLA.

suitable. In tbe calcium oxalate precipitation it was found
necessary, however, to prolong tbe time of precipitation to 24
hours in order that uniform values could be obtained. The deter-
minations of phosphorus were carried out photometrically by
Fiske-Subbarows method. Traces of albumin were removed
prior to phosphor determinations with 7 % trichloroacetic acid.

Results.

I. Uptake and retention of calcium and phosphorus
in normal rat bones.

Before entering upon the actual vitamin experiments it seemed
advisable to investigate in general the behaviour of normal bone
in the uptake and retention of calcium and phosphorus, a problem
which, in view of the renewal of bone, is not devoid of interest
but which has not been more closely investigated with chemical
methods.

L VftaTce of calcium.

The following table summarises the results of six experiments.
In three of them was studied the calcium absorption of fore leg
preparations of varying weight and in the other three that of the
hind leg preparations.

Table 1.

Bone preparation

Weight of bone
preparation, mg

Calcium
uptake, mg

Fore leg

500

0.12

600

0.08

890

0.06

Hind leg

950

0.17

1,150

O.IO

1,150

O.OG

It can be seen from the table that the normal rat bone has the
ability to take up calcium from the incubation solution. The
analytical differences are slight but, nevertheless, convincingly
outside the limits of error of calcium analyses.

STUDIES ON THE ROLE OP VITAMIN D.

289

2. Rclcntioii of calcium.

Tlie following table gives tlie results of e^qperiments performed
Avith the same amount of bone preparations from bones of approxi-
mately the same weight as above, with the exception that cal-
cium-free physiological sodium chloride solution was used as
incubation solution-

Table 2.

Bone preparation

Weight of bone
preparation, mg

Loss of
Cnlcinm, mg

Fore leg

550

0.40

570

0.62

> *

650

0.65

Hind kg

1,100

0.G6

1,300

0.68

1,450

O.Gl

The table shows that a distinct excretion of calcium occurs
from the normal bones into the surrounding solution. Regarding
the weight of the preparations it is evident that the fore leg
preparations have given pronouncedly largest amounts calcium
per weight unit into the solution.

3. UplaJce of phosphorus.

In the experiments concerning the uptake of phosphorus an
incubation solution containing 10 mg % phosphorus was first
used. No uptake of phosphorus from the solution could be observed
here as shown by the values of the four experiments recorded in
table 3; on the contrary the phosphorus content of the incubation
solution was higher after the incubation than before it. In later
experiments (in collaboration with Antell and Bardy) higher
concentrations of phosphorus up to 100 mg % were used. Nev-
ertheless, even in the highest concentrations no decrease was
noted in the phosphorus content of the incubation solution. On
the contrary, phosphorus seemed to be excreted, as in the above
mentioned phosphorus experiments, from the bone into the
incubation solution.

290 V. V. KRAEMER, B. LAKDTMAN AND P. E. SIMOLA.

Table 8.

Bone preparation

TVeiglit of bone
preparation, mg

Phosphorns
uptake, mg

Fore leg

475

-0.19

550

-0.86

Hind leg

1,100

- 0.12

1,150

— 0.81

4. Retention of phosphorus.

In Table 4 are compiled tbe results of experiments performed
•with preparations from fore and bind legs under equal conditions
as those above concerning tbe calcium retention using physiological
sodium chloride as the incubation solution.

Table 4.

Bone preparation

"Weight of bone
preparation, mg

Loss of

Phosphorns, mg

Fore leg

550

0.26

570

0.88

> .

650

0.38

Hind leg

1,100

0.62

1,300

0.38

1,400

0.86

It can be seen that a pronounced transfer of phosphorus from
the bone preparations into the incubation solution occurs in all
cases. It is difficult to conclude from these few preliminary experi-
ments which factors cause and influence this transfer of phos-
phorus. Apart from the pmely physical factors there is a possi-
bility that the transfer of phosphorus into the solution is in some
relation to the activity of the bone phosphatases. We are going
to examine later in detail the excretion of phosphorus into the
solution.

In any case, the preliminary studies in vitro with normal bone
preparations showed that between the bones and the solution an
exchange of calcium and phosphorus occurs which is distinctly

STUDIES ON THE EOLE OF VITAMIN D. 291

demonstrable by chemical methods. The technique employed
makes possible the investigation of the calcium and phosphorus
metabolism in the bones under different conditions.

IL Uptake and retention of calcium and phosphorus
in the D-hypervitamlnosis.

1. Uptake of calcium.

Table 5 summarises the results of parallel experiments, con-
cerning the calcium uptake, carried out "with fore leg preparations
of D-hyper\'itaminotic and control animals. The hypervitaminotic
animals had received, during three days, 24,000 I. U. daily, the
control animals receiving sesame oil only.

Table 5.

Bone preparation

IVeight of bone
preparation, mg

Calcium
uptake, mg

Controls

Fore leg

600

0.08

625

0.08

> ,

720

0.06

720

0.02

Hyperritaminotic animals j

500

0.31

640

0.16

672

0.19

680

0.19

A comparison of the calcifying power of bone preparations of
approximately the same weight from D-hypervitaminotic and
control animals shows that liberal admimstration of vitamin D
has distinctly increased the ability of the hones to absorb calcium
from the solution. The mean value for calcium absorption in the
normal cases is only about one-fourth of the corresponding value
in the hypervitaminotic cases. The experiments support the view
that vitamin D somewhat makes the bones “more hungry for
calcium than usual.

292

V. V. KRAEMER, B. LANDTMAN AND P. E. SIMOLA.

2. Retention of calcium.

The influence of vitamin D on the uptake of calcium led to
tlie assumption that tHs vitamin also increases tke ability of the
bones to retain calcium. In order to elucidate this point a series
of experiments were made, the results of which are recorded in
Table 6. The production of hypervitaminosis in these cases was
accomplished by administering daily 12,000 I, U, of vitamin D
during four days. The experimental material in the incubation
experiments consisted of hind leg preparations.

Table 6.

Bone preparation

"Weight of bone
preparation, mg

Loss of
Calcium, mg

Controls

Hind leg

1,090

0.B6

> . «

1,100

0.66

» . . . .

1,195

0.44

> . .

1,300

0.58

Hypervitaminotic animals

1,205

0.44

1,320

0.60

1,370

0.68

1,650

O.BO

A comparison of the results in both groups shows no such
difference in the loss of calcium as in the above experiments on
the uptake of calcium. It is difficult to draw conclusions from the
results, as the weight of the bone preparations used in the control
experiments was, on the average, lower than in the experiments on
D-hypervitaminosis. The variations in both groups are, however,
so great that the final solution of the problem seems possible only
by means of an extensive test material. (The supplementing
experiments carried out so far, showed in some cases a distinct
tendency for a decrease of calcium excretion in D hypervita-
minosis.)

STUDIES ON THE ROLE OP VITASIIN D,

293

3. Retention of pJjospJtonts.

Table 7 gives data from experiments concerning the loss of
phosphorus Tvlieii hind leg preparations of D-hypervitaminotic
animals and the controls were kept in phosphorusfree sodium
chloride solution. In these cases vitamin D was administered in
daily doses of 12,000 I. U. during four days.

Table 7.

Bone preparation

Weight of bone
preparation, mg

Loss of

Phosphorns, mg

Controls

Hind leg

1,090

0.84

1,100

0.B2

1,195

0.38

1,300

0.88

Hypemlaminotic animals

1,205

0.81

1,325

0.29

1,370

0.29

1,650

0.24

It can be seen from the table that a difference exists be-
tween the values for phosphorus excretion in D-hypervitamino-
sis and controls, the excretion being lower in the former group.
Accordingly vitamin D would somehow be able to effect a retention
of inorganic phosphorus in the bone. It may be remarked against
the arrangement of the experiments that, due to the different
weight of the bones, the groups are not fully comparable to each
another. Considering, however, also the above-reported values for
the retention of phosphorus in the control animals (Table 4), the
values for excretion lead to the conclusion that the effect is
ascribable to vitamin D.

TTT. Uptake and retention of calcium and phosphorus

in rachitis.

following the above observations on D-hypervitaminosis, our
chief attention was directed to the possible effects caused by D-

294

V. V. KRAEJIER, B. LANDTMAN AND P. E. SISIOLA.

hypovitaminosis under the same experimental conditions. The
rats employed in the experiments showed typical symptoms of
rachitis as a result of prolonged chronic deficiency of vitamin D.
By chemical examination a pronounced decrease of inorganic
phosphorus was discernahle in the serum.

1. UptaJce of calcium.

Table 8 gives the values of calcium uptake obtained with the
hind leg preparations of two rachitic and two normal rats. It
is noteworthy that neither of the hind leg preparations of the one
rachitic rat has taken up calcium from the solution, on the con-
trary, they have given off calcium into the solution. Thus the
phenomenon concerned would be completely contrary to that in
D-hypervitaminosis in which calcium uptake was increased. The
changes in the calcium content of the solution caused by the
preparations of the other rachitic rat are, however, in accordance
with those in the control animals. Bor the final settlement of
this point more experiments with a more extensive material are
naturally needed.

Table 8.

Bone preparation

Weight of bone
preparation, mg

Calcium
uptake, mg

Controls

Fore leg

470

0.14

460

0.18

700

0.02

> ...

750

0.02

Rachitic animals

620

— 0.13

650

-0.22

> ...

450

0.11

450

0.13

2. Retention of calcium.

A comparison of the values of calcium uptake, obtained with
the bones of rachitic rats (Table 9) and the corresponding values
of normal animals, with those of the previous experiments on D-

STUBEES OK THE ROLE OP VITAMIN D. 295

liypervitaminosis show that there is an apparent tendency to-
wards reduced ability to retain calcium in the rachitic cases.
This result is again contrary to what was the case in D-hyper-
A-itaininosis.

Table 9.

I Bone preparation

Weight of bono
preparation, mg

Loss of
Calcinm, mg

{

Controls

1 Fore leg

340

0.48

I >

300

0.45

500

0.47

i *

470

0.93 (?)

Rachitic animals

’

380

0.G2

> .

360

0.6 0

’

300

0.96

320

0.97

3. Jlcicntion of fJiosphorits.

It is evident from Table 10, which gives the values obtained
with bone preparations of rachitic and control animals, that no
distinct difference can be observed between the two groups.

Table 10.

Bono preparation

Weight of hone
preparation, mg

Loss of

Phosphorns, mg

Controls

Fore leg

380

0.S8

420

0.27

500

0.24

470

0.24

Rachitic

animals

380

0.27

360

0.30

300

0.27

320

0.20

296 V. V. KBAEMER, B. LANDTMAN AND P. E. SIMOLA.

Direct comparison of the results is rendered difficult by the fact
that the bones of the control animals have, on the average, been
bigger than those of the rachitic animals. If the amount of
excreted phosphorus is calculated per weight unit of bones, it
appears that in rachitis the amormt of excreted phosphorus per
weight unit of bone is somewhat greater than in the controls.

Discnssion of the results.

The preliminary experiments which were carried out in order
to obtain a basis for subsequent vitamin experiments, brought into
light facts that are not devoid of interest in view of the mineral
metabolism of the bone. It appeared from the experiments that
the bone is able to take up calcium from the surrounding solution
containing calcium. On the other hand, if the bone is incubated in
the calciumfree solution it gives up calcium into the solution.
With regard to inorganic phosphorus, the bone did not take it up
although the phosphate concentration was raised high. On the
other hand, when the bone was incubated in solutions with or
without phosphorus, inorganic phosphorus always appeared.

An explanation to these variations is naturally offered by
the simple physical diffusion. On the other hand, the possibility
is not excluded that some active function of cells plays a part- in
these phenomena. These points are further investigated in new
experiments now in progress. It must be mentioned in this con-
nection that the experiments of Eobison on the ossification had
proved that the bone does not take up phosphorus in inorganic
form, but as phosphoric acid esters.

Our experiments showed that D-hypervitaminosis is charac-
terised by an increased ability of the bones to uptake calcium
when the incubation takes place in solutions containing calcium,
and by a decreased tendency to give up calcium into the surround-
ing Ca-free solution. At the same time, the excretion of inorganic
phosphorus into the phosphorus-free incubation solution showed a
tendency to decrease. On the other hand, in a deficiency of vita-
min D a decrease was noted in the uptake of calcium, and an in-
creased tendency towards an excretion of calcium into the cal-
ciumfree incubation solution. Through the effect of vitamin D
the bone would accordingly become somehow more hungry for
calcium and also attain an increased ability to retain calcium and
possibly also phosphorus.

STUDIES ON THE ROLE OF VITAMIN D. 297

It was mentioned already that the way of action of vitamin
D on the bone — whether direct or indirect — has not been
definitely settled up till present. It has been assumed that the
changes in the bone, specific to rachitis, are merely consequences
of the disturbances in the calcium and phosphorus of the serum
somehow caused by vitamin D. The observations described above
indicate that vitamin D has a direct influence on the bone.^ This
does not mean, that the effect of vitamin D could not appear in
other tissues. It is conceivable that the ability of the cells in
general — or at any rate of certain groups of cells — to take up
calcium or retain it, is somehow dependent on vitamin D. Some
experiments with other tissues, carried out in connection with
this work have to some extent revealed that also in other tissues
it is possible to bring about phenomena similar to those described
above. That the resorption of calcium from the intestine is dis-
turbed by vitamin D deficiency has been earlier established, as
mentioned above.

Summary.

The object of the work has been to investigate, by means of
experiments in vitro, the effect of vitamin D on calcium and
phosphorus metabolism in rat.

The normal bone has the ability to take up calcium from the
incubation solution under certain conditions and, on the other
hand, to give up calcium into the calcium-free incubation solution.

The bone was not able to take up inorganic phosphorus in the
incubation experiments. On the other hand the excretion of in-
organic phosphorus into the solution could always be demon-
strated — both in incubation solutions with and without phos-
phorus.

The bones of rats which had received large amounts of vitamin
D seemed to be able to take up relatively more calcium than the
bones of normal rats. On the other hand it is possible that the
bone is then to some extent more than normally able to take up
calcium and phosphorus.

With regard to the deficiency of vitamin O, a tendency towards
a decreasing uptake of calcium and increased loss of calcium was
noted.

The experiments support the view that vitamin D has a direct
influence upon the mineral metabolism of the bone.

298

V. V. KRAEMER, B. LAMTMAN AND P. E. SIMOLA.

Eeferenees.

Fleischsiann, C. F., Arcli. Zellforscli. 1937, 19 . 1031.

Harris, L. J., Lancet 1932. 222 . 1031.

Nicolaysen, R., Biocliem. J., 1937. 31 . 107. 122. 323. 1086.
Robison, R., and K. M. Soames, Ibidem 1930. 24 :. 1922.

— , and A. H. Rosenheim, Ibidem 1934, 30 . 684.

Rosenheim, A. H., Ibidem 1934, 30 . 708.

Shipley, P. G., B, Kr amer and J. Howland, Amer. J. Dis. Children
1925. 30 . 37.

— , — , — , Biochem. J., 1926. 24 . 379.

Aus der Anatomischen und der Pharmakologischen Abteilung
des Karolinischen Instituts in Stockholm.

tiber den Eiiifliiss lokaler Haiitscliadigiingen
(nieclianisclie Yerletziing, Erfrierung, Verbreii-
iiimg) auf die periphere Blntyerteilimg.^

Yon

GOSTA von REIS und PRITIOF SJOSTRAND.

(Mit 2 Figuren im Tnit.)

In friilieren Arbeiten (G. von Reis und R. Sjostrand 1937,
1938, 1940) sind die Resultate von Untersucliungen iiber den Ein-
fluss verschiedener chemischer und physikalisclier Hautreize auf
die Blutmenge in peripbcren Blutgefassen der Leber und Nieren-
rinde veroffentlicbt worden. Bei melireren Reizmitteln war dabei
eine erliebliclie Zunahme der Blutmenge in diesen Gefassen kon-
statiert worden.

Bei einer ersten Analyse des Zustandekommens dieses Effekts
hat sicb derselbe als von einer intakten Hautinnervation abbangig
erwiesen. Irgendein Einfluss durch. Resorption von Substanzen
aus der Haut, welche auf die peripheren Blutgefasse einwirken
konnten, hat sich dagegen nicht feststellen lassen.

Die Absicht bei den in der vorliegenden Arbeit beschriebenen
Versuchen war in erster Linie zu untersuchen, inwieweit eine lo-
kale Hautschadigung durch z. B. Verbrennung, Erfrierung oder
mechanische Einwirkung einen Einfluss auf die peripheren Blut-
gefasse in inneren Organen ausubt, von welchen bisher Leber und
Rierenrinde, bei Verbrennung auch der M. masseter, untersucht
wurden. Hierdurch whrde die Anschauung ziemlich direkt ge-
stiitzt werden, dass diese Blutgefasse prirnar einen beitragenden
Raktor bei der Entstehung des Schocks nach ausgedehnteren
derartigen Schadigungen darstellen.

Der Redaktion am 14. September 1940 zugcgangen.

21 — i01323. Acta phys. Scandmav. VoI.I.

300 QOSTA VON REIS UND FRITIOF SJOSTRAND.

Ausserdem suchten -wir melir allgemein zu entsclieiden, ob bei
einer Warmescbadigung in der Haut Substanzen gebildet oder
fieigemacbt werden konnen, vrelche nacb. ibrer Resorption im-
stande sind, auf die peripberen Blutgefasse im Sinne dex Annabme
von Leavis (1927) einzuvdrken. Zu diesem Zweck haben wir bei
lokalen Hautscbadigungen den Effekt, welcber an den peripberen
Blutgefassen resultiert, wenn die Scbadigung Haut mit intakter
Innervation trifft, mit demjenigen einer entsprecbenden Scha-
digung von denervierter Haut verglicben.

Wabrend uber die verscbiedenen Wirkungen der Verbrennung
ein umfangreicbes Scbrifttum entstanden ist sind Angaben liber
den Effekt von Erfrierungen nur sparlicb vertreten, und, soweit
wir finden konnten, experimentelle Arbeiten liber die diesbezlig-
lichen Folgen mechaniscber Hautscbadigungen liberbaupt nicbt
vorbanden. Hinsicbtlicb der Verbrennung wird auf Monograpbien
und ijbersiebtsreferate von Marchanb (1908), Stockis (1913).
Wilson (1929) und Harkins (1938) verwiesen.

Es ist vor allem die Ursacbe des in engem Zusammenbang mit
ausgedebnten Brandwunden auftretenden Scbocks, welcbe man
zu analysieren sucbte, und es bandelte sicb dabei im grossen ganzen
darum, zu entscbeiden, ob dieser durcb eine Einwirkung auf das
Nervensystem zustandekommt ~ welcbe entweder eine Hem-
mung innerhalb verscbiedener vegetativer Zentra in der Medulla
oblongata oder eine libermassige Reizung der Hebennieren ber-
beifiibren konnte, durcb die der Adrenalin vorrat dieses Organs
erscbopft wurde, mit darauf folgender allgemeiner Gefassdilata-
tion — , durcb eine Vergiftung mit Substanzen aus dem gescba-
digten Gewebsgebiet, dutch Plasmaverlust infolge von Odemen,
oder scbliesslicb durcb Bakterienwirkung.

Mebrere Autoren (Dohrn 1901, Wilms 1901, Stockis 1903)
haben die Ansicht geaussert, dass der Schock wahrsoheinlich
einem Zusammenwirken mehrerer Eaktoren zuzuschreiben sei.
Heutzutage scheint man jedocb allgemein eher einen Effekt dutch
Intoxikation als den dominierenden Eaktor anzunehmen, wenn
sicb aucb eine sichere Stlitze flir diese Erklarungsweise nicht
anflihren lasst;

Seitdem Lewis (1927) und Krogh (1929) auf die Moglicbkeit
hingewiesen baben, dass der Schock dutch Histamin- oder H-
Substanzvergiftung zustandekommen sollte, glaubten mehrere
Forscher, experimentell gewisse Anbaltspunkte flit diese Hypo-
tbese finden zu konnen, wabrend andere betracbtliche Unter-

EINFLUSS LOKALER HAUTSCHADIGUKQEN. 301

schiede zwischen Histamin- und Yerbrennungsscliock fest-
stellten.

Die A^ersuche, den A^’erbrennungsacliock mit- dem Histamin-
schock gleicbzusetzen, sind ein Ausdruck der Auffassung des er-
steren als Folgc einer Kreislanfsinsiiffizienz, n. a. anf Grand einer
hocbgradigen Dilatation der peripheren Blutgefasse.

Auf das A'orliegen einer solclien Dilatation hat man u. a. aus
Blutdruckversuchen geschlossen (Sonnenbubg 1878, Stockis
1903 u. a. m.). Simo^’aet (1930) beobachtete ausserdem die Farbe
der Organe, nachdem die A^ersuchstiere durch A^'erblutung an ver-
schiedenen Zeitpunkten nach einer in Athernarkose liervorgeru-
fenen ausgedehnten A’’erbrennung (-/a der Korperoberflache)
getotet worden waren. Er fand dabei, dass vor allem die Bauch-
organe — besonders die Eieren — bei den Tieren mit Brand-
vrunden roter •vraren als bei den Kontrolltieren. Bei in Einzelfallen
vorgenommener mikroskopischer Untersuchung sail S., dass die
Kapillaren stark erweitert waren.

Die Allgemeinwirkung von Erfrierung und Auftauung eines
Gewebsgebiets ist nor in einer geringen Anzahl experimenteller
Aibeiten behandelt worden. Peeiffer (1927) gibt an, dass der
Effekt dieser Art von Gewebsschadigung dem bei Verbrenmmg
entspreche und auf Desorption von Zerfallsprodukten aus dem
geschadigten Gewebsgebiet beruhe. Experimentell wird diese
Behauptung nicht gestiitzt.

Dagegen hat Harkins gemeinsam mit verschiedenen Mitar-
beitfern (1934, 1935, 1937) nachgewiesen, dass man bei Hunden
durch. Kalteschadigung einen Schockzustand hervorrufen kann,
welcher dem bei A^erbrennung in mehreren Beziehungen iihnlich
ist (Bluteindickung durch Plasmaverlust, herabgesetztes Blu-
tungsvolumen). AYie dieser Schockzustand ausgelost wird scheint
nicht Gegenstand der Untersuchung gewesen zu sein, obwohl
man hierin ein gewisses prinzipielles Interesse erblicken konnto.

Methodik.

Da die allgemeine A^ersuchsanordnung bei diesen A'^ersuchen ganzlich
mit derjenigen iiberemstimmt, iiber welche wir in imseren friiheren
Arbeiten berichtet haben, verweisen wir auf diese (1938, 1940).

Als Versucbstiere haben wir Meerschweinchen mit einem Gewicht
von 400 — 600 g verwendet. Tiere beider Geschlechter fanden in unge-
fahr gleichem. Masse Verwendung, und eine an das Geschlecht gebim-
dene Variation wurde nicht beobachtet.

302 GOSTA VON KEIS UKD FBITIOP SJOSTRAND.

Das beliandelte Hautgebiet -war 3 x 6 bis 4 x 6 cm gross und er-
streckte sich quer iiber die Bauchhaut; die kaudale Grenze lag in der
Hohe der Nabelebene.

Die Versucbe wurden in Pernoctoimarkose vorgenommen.

Die Tiere wurden durcb Zerquetschung des Halsmarks getotet.

Die Blutmenge in den peripberen Blutgefassen wurde nacb T. Sjo-
STKAND (1934) und von Eeis, Silfverskiold, F. Sjostrand und T.
Sjostranb (1938) bestimmt, d. b. die Anzabl roter Blutkorpercben pro
ram® Gcwebe in der Leber und Nierenrinde wurde nacb selektiver
Farbung von Mikrotomscbnitten durcb Vergleicb unter dem Vergleicbs-
mikroskop rait einer Serie von Sfcandardpraparaten festgestellt. Bei
Verbrennungsversucben vmrde ausserdem die Anzabl offener Kapillaren
pro mm® des Querschnitts durcb den M. masseter durcb Zablung be-
stimmt.

Eine obetflacblicbe Verbrennung vrurde auf zwei vetscbiedene Arten
bervorgerufen, teils durcb Brennen mit einem O.i mm dicken Messing-
blecb oder einem 0.05 mm dicken Kupferblecb, welcbes zum Gliiben
gebracht und dann zwei- bis dreimal 2-^ Sek. gegen die Haut gepresst
vrurde, teils durcb intensive Bestrablung mit einer Warmelampe. Durcb
die geringe Warmekapazitat der diinnen Metallblecbe vrurde die Ver-
brennung sebr oberflacblicb. Die von uns verwendete Warmelampe
{»61ory«) sendet Warmestrahlen von sowohl hoherer ■wie niedrigerer
Wellenlange aus. Die Tiere mirden mit derselben 5 — 10 Min. intensiv
bestrablt, vrobei eine deutlicbe Verbrennung zustandekam.

In denjenigen Fallen, vro die Verbrennung mittels Metallblecbs her-
vorgerufen vrorden war, wurden die Tiere 45 Min. nacb dem Eingriff
getotet, wabrend sie bei den Versuchen mit der Warmelampe 35-^0
Jlin. nacbdem die Hautveranderungen sicb entwickeln konnten getotet
wurden.

Eine lokale Kaltescbadigung wurde entweder durcb Bespritzen mit
Chloratbyl erzeugt oder dadurcb, dass ein mit Koblensaurescbnee
gefiillter Kupferbebalter von geeigneter Form gegen die Haut gepresst
^'rurde.

Die Haut wurde in beiden Fallen 10 — ^20 blin. in gefrorenem Zustand
gehalten, wonacb man sie auftauen liess. Ausserdem land partielle
Auftauung wabrend der Bespritzung mit Chloratbyl und der Gefrierung
mit Koblensaurescbnee statt, da es auf andere Weise unmoglich war,
das Gefrieren auf die Haut zu beschranken.

Die Tiere wurden 30 — 15 Min. nacb dem Auftauen getotet.

Eine mechaniscbe Hautverletzung wurde entweder dadurcb erzielt,
dass die Haut mit einer Rasierklinge zerschnitten wurde oder durcb
Zerkratzen mit einer feinen Sage. Die Tiere wurden 45 Min. nacb dem
Eingriff getotet.

In denjenigen Fallen, wo wir eineu deutlichen Effekt erbielten, wurde
ein entsprechender Versuch an Tieren ausgefubrt, deren Bauchhaut
ihrer segmentalen Innervation beraubt worden war, nacb derselben
Methode, wie sie in friiheren Arbeiten (1938) angegeben ist.

EINFLUSS LOKALER HADTSCHADIGUKGEN.

303

Karze Besclireibung der crzielten HaiitTerfinderimgen.

Beim Brenncn mit dem Messingblech wurde die Haut stark
zusammcngezogen und verdickt. Sie tvurde hart und unelastisch,
blasste ab und nahm teilvreise eine braunliche Fiirbung an. Von
der Innenseite gesehen hat ein abprapariertes Hautstiick eine
Starke Gefassinjektion aufgewiesen. Ein Odem im Corium wurde
nicht beobachtet.

Die Haut war teilweise mit der darunterliegenden Muskulatur
adharent. Die Epidermis war etvras odematos-verdickt, und die
aussere Muskelschicht war entsprechend der Stelle der Verbren-
nung in geringerer Ausdehnung etwas gerotet und odematos.
Peritoneum und Diirrae waren unversehrt.

Die mit dem diinnen Kupferblech hervorgerufene Brandwunde
war von geringerer Ausdehnung. Die Haut war auch hier, wenn
auch in gcringerem Grade, zusammengezogen und ihrer Farbe
und Konsistenz nnch verandert, ohne dass es jedoch, abgesehen
von vereinzelten Fallen, zu einer Braunung gekommen ware. Die
Haut Hess sich Icicht von der Untcrlage ablosen, und eine Ein-
wrkung auf die darunterliegende Muskulatur wurde nicht beob-
achtet. In einzelnen Fallen sah man ein uncihebliches Odem
im XJntcrhautzellgewebe.

Beim Brennen mit der 'Warmelampe wurde die Haut zunachst
gerotet, worauf sich innerhalb eines Bezirks mit nahezu kreis-
runder Begrenzung und einem Durchmesser von 4 — 5 cm ein
Odem in der Haut bUdete. Im Zentrum dieses ddematosen Gebiets
blasste spater die Haut ab, und das Odem bildete einen Wall rings
um die so entstandene abgeblasste Zone, welche einen Durch-
messer von etwa 2 cm hatte. Im ubrigen ahnelten die Veran-
derungen denjenigen beim Brennen nrit Metallblecb, wenn sie
auch mehr in die Tiefc gingen, sodass die Muskulatur in grosserem
Ausmass geschadigt wurde.

Bei keinem Fall wurde Blasenbildung in der Haut beobachtet.

Die Odeme waren von unerheblicher Ausdehnung, sodass ein
Einfluss auf den Fliissigkeitsgehalt des Blutes oder der Gewebe
keine Fehlerquelle von Bedeutung bilden konnte.

Bei der Gefrierung der Haut mit Chlorathyl entstanden Ver-
anderungen, welche in gewissem Masse denjenigen beim Brennen
mit der Warmelampe ahnlicli waren. Zentral bildete sich also
ein blasser Hautbezirk mit einer schwach-braunlichen Verfarbung,

304

GOSTA VON REIS UND FRITIOF SJOSTRARD.

I L eier ^ AnzM BlutkSrperchen ii?

'NierenrindB faujendet7 pro mrrA Go\»^h‘^

Die Pfeile nnter der Abszisse bezeicbnen die
Stan dardpraparaten .

Diagramm 1.

"Werte bei den verwendeten

umgeben. von einem weniger stark ausgepragten Odemwall. Die
Haut vat mit Ausnabme der anamischen Zone in dem verletzten
Gebiet gerotet.

Nacb Gefrierung mit gekiibltem Metall -wurde nur eine Rotung
der Hant beobacbtet.

Bei der mecbaniscben Hautverletzung mittels Rasierklinge

EINFLUSS LOKALEE HAUTSCHADIGUKGEN.

305

Die Pfeile unter der Abszisse bezeichncn die Werte bei den verwendeten

StandardprSparaten;

Diagramm 2.

wurden nur oberflachliclie kutane, dichtliegende Solinitte erzeugt.
Die mit der Sage hervorgerufenen Verletzungen waren ebenialls
oberflacblicb und bescbxiinkten sich auf Epidermis und Oorium.
In beiden Eallen kam es nur zu unbedeutendex Blutrmg.

Hinsichtlicb der statistiscken und grapbiscben- Bearbeitung
der erbaltenen Werte veru-eisen w auf unsere friibere Arbeit
iiber den Einfluss lokaler physikaliscker und cbemiscber Haut-
reize auf die peripbere Blutverteilung (1940).

306

GOSTA VON EEIS UND FEITIOF SJOSTEAND.

Yersuchsergebnisse.

KontroUtiere.

Da wir mit denselben Versucbsbedingungen gearbeitet baben,
wie sie von uns in friiberen Arbeiten beschrieben sind, gelten die
in diesen veroffentlicbten Kontrolltierwerte fiir die peripbere
Blutmenge in der Leber nnd Nierenrinde sowie fiir die Anzabl
offener KapUlaren im Querscbnitt des M. masseter aucb in diesem
Zusammenbang.

Die Mttel der Werte fiir die peripbere Blutmenge betrugen bei
der Leber (49 Tiere) M = 268,000 und bei der Nierenrinde (44 Tiere)
M = 363,000 Blutkorpercben pro mm® Gewebe. Diagramm 1 A
lasst die Streuung der Primarwerte erkennen. Die Anzabl offener
KApUlaren im Querscbnitt des M. masseter war im blittel (19
Tiere) 1,330 pro mm®.

Die Primarwerte fiir die Anzabl offener Kapillaren im M. masse-
ter von Kontrolltieren findet man in unserer friibaren Arbeit
liber den Einfluss lokaler cbemiscber und pbysikaliscber Haut-
reize auf die peripbere Blutverteilung (1940).

Lokale Verbrennung.

Bei Versucben mit Brennen mittels Messingblecbs wurden die
Mittelwerte (17 Tiere) fiir die Leber M = ,559,000 und fiir die
Nierenrinde SI = 253,000 Blutkorpercben pro mm® Gewebe. Aus
Diagramm 1 B wird die Streuung der Primarwerte ersicbtlicb.
Im SI. masseter erbielten wir imSIittel (15 Tiere) 1,650 blutgefiillte
Kapillaren pro mm® Querscbnitt. Die Primarv^erte findet man in
Tab. 1.

Aus diesen Werten gebt bervor, dass die peripbere Blutmenge
in der Leber auf das doppelte der entsprecbenden Kontrolltier-
werte gestiegen ist. In der Nierenrinde dagegen ging diese Blut-
menge um 30 % zuriick, eine Verringerung, welcbe jedoch fiir dieses
Organ innerbalb der Eeblergrenzen der Sletbode liegen diirfte.

Die Anzabl offener Kapillaren pro Flacbeneinbeit im Quer-
schnitt des SI. masseter nabm um 25 % zu. Die Differenz der
mittleren Eebler (t) zwuscben dieser Serie und der Kontrolltier-
serie betriigt 8.4, wesbalb die Differenz der beiden Slittelwerte
statistiscb gesicbert ist.

EINFLUSS LOKALER HAUTSCHADIQUNGEN. 307

Tnljelle 1.

Anzahl offener Kapillaren pro mm~ Querschniti des M. masseter 45
Min. nach Erzeugung einer oberflachlichen loJcalen Verbrennung.

Isr.

intakte Hunt
innervation

denervierte

Haut

1,690 ± 53

1,330 + 46

1,450 + 47

1,340 ± 47

1,630 ± 52

1,430 ± 44

1,630 + 34

1,180 ± 34

1,700 ± 37

1,520 ± 49

1,670 + 44

1,250 + 29

1,630 ± 43

1,430 ± 36

1,680 ± 39

1,300 ± 40

1,710 +47

1,310 + 28

1,600 ± 45

1,520 ± 53

1,770 ± 45

1,450 + 44

1,610 + 37

1,700 ± 51

1,760 ± 48

1,570 + 39

M-

1,650

1,370

Bei durch Bestrahlung mit der Wiirmelampe erzielter Ver-
breanung erbielten wir im Mifctel (10 Tiere) in der Leber 430,000
und in der Nierenrinde 294,000 Blutkorperchen pro mm® Gewebe.
Diagramm 1 C zeigt die Streuung der Primarwerte.

Die periphere Blutmenge ist in diesem Fall in der Leber auf
das reichlicb l^/afacbe des entsprechenden Kontrolltierwerts
gestiegen, aber bezuglicb der Nierenrinde lasst sicb auch bier
kein sicberer Einfluss konstatieren.

Um festzustellen, ob das Leberparencbym beim Brennen eine
Erwarmung erfubr, warden wabrend des Eingriffs Temperatur-
bestimmungen an verscbiedenen Stellen in der Baucbboble vorge-
nommen. Die Messungen warden mit einem empfindlicben
Qaecksilbertbermometer aasgefiibrt.

Wenn die Verbrennang anterbalb des kandalen Leberrandes
liegende Haatbezirke trifft "nird die Erwarmang sebr anerbeblicb.
Nnr zwiscben Bancbwand und Leberparencbym, an dem verbrann-
ten Hautgebiet unmittelbar benacbbarten Stellen, Hess sicb eine

308 QOSTA VON BEIS UND FRITIOF SJOSTRAND.

Temperatursteigerung von TO nachweisen, Tiefer in der Bauch-
hohle und zwxsclien den Leberlappen. Melt sich die Temperatnr
somit konstant.

Ware der Effekt einer ortlicken Verbrennnng anf die peripberen
Blutgefasse der Leber eine Eolge von Erwarmung des Leber-
parencbyms, dann wiirde derselbe innerbalb verscMedener Teile
der Leber verscMeden stark ausgepragt sein. Ein solcbes Verbalten
vnirde indessen niemals beobacbtet, vresbalb der Effekt anf an-
dere Weise zustandekommen diirfte.

Lokale Erfrierung.

An 10 Tieren -wurden lokale Erfrierungen durcb Bespritzen mit
Cbloratbyl und bei 7 Tieren mittels geldiblten Metalls, welcbes
gegen die Haut gepresst wurde, hervorgerufen. Da die Eesultate
dieser beiden Gefriermetboden ubereinstimmen wurden samt-
licbe Versucbe in eine Serie zusammengefasst.

Die Mittelwerte der peripberen Blutmenge in der Leber und
Nierenrinde wurden 510,000 bzw. 264,000 Blutkorpercben pro
mm* Gewebe. Diagramm 2 D veranscbaulicbt die Streuung der
Primarwerte.

Aus diesen Werten wird ersicbtlicb, dass, verglicben mit den
entsprecbenden Kontrolltierwerten, die peripbere Blutmenge in
der Leber im Mittel verdoppelt wurde, wabrend sie in der Nieren-
rinde um 27 % sank.

Lokale mechanisclie Hautverletzung.

Die beiden Metboden zur Erzeugung einer mecbaniscben Haut-
verletzung gaben identiscbe Resultate, wesbalb die Versucbe zu
einer Serie vereinigt wurden. Die Mittelwerte der Bestimmungen
an 9 Tieren betrugen fvir die Leber 359,000 und fiir die Nierenrinde
227,000 Blutkorpercben pro mm^ Gewebe. Die Anordnung der
Primarwerte gebt aus Diagramm 2 E bervor.

Wabrend sicb also kein sicberer Effekt auf peripbere Blutge-
fasse in der Leber konstatieren liess ist die entsprecbende Blut-
menge in der Nierenrinde um fast 40 % zuriickgegangen.

Lokale Sckfidigung denervierter Haut.

Bei lokaler Verbrennung von denervierter Haut, wobei mit dem
Messingblecb gebrannt wurde, betrugen die Mittelwerte (11 Tiere)

EINFIiUSS LOKALER HADTSCHADIGUEGEN.

309

fur die periphere Blutmenge in der Leber 535,000 und in der Nie-
renrinde 273,000 Blutkorpercben pro mm® Gewebe, gegeniiber
559,000 bzw. 253,000 bei entsprechender Behandlung von Tieren
mit intakter Hautinnervation. tjber Streuung der Primarwerte
siebe Diagramm 2 F. Die Bestimmungen der Anzabl offener
Kapillaren im Masseterquerscbnitt an diesen Tieren (Tab. 1)
ergaben im Slittel (11 Tiere) 1,370 gegeniiber 1,650 bei gebrann-
ten Tieren mit intakter Hautinnervation. Die Differenz der mitt-
leren Felder bei diesen Serien betragt 7.6, -vvesbalbdie Differenz
der Jlittelwerte statistisch siclier ist. Der Wert 1,370 stellt kei-
nen statistisclien Unterschied von dem Wert fiir die Kontrolltiere
dar.

Aus diesen Werten ergibt sick, dass der Effekt auf peripbere
Blutgefasse in der Leber bei ortlicber Verbrennung der Baucb-
baut aucb dann resultiert, wenn die Haut und teilweise die
Baucbmuskulatur vor dem Brennen ibrer segmentalen Innerva-
tion beraubt vorden war. Dagegen bleibt der Effekt auf die
Kapillaren im M. masseter in denjenigen Fallen aus, wo die ver-
brannte Haut denerviert war, wesbalb diese Wirkung von einer
intakten Hautinnervation abbangig sein muss. In keinem der
Fade wurde ein sicherer Effekt auf die entsprechenden Blutgefasse
in der Nierenrinde erzielt.

Um bei diesen Versucben mit Sicberbeit ausscbliessen zu kon-
nen, dass der Effekt von der Innervation des gescbadigten Gewebs-
gebiets abbangt, muss man natiirlicb voraussetzen, dass nicbt
undener\derte Gewebspartien, beispielsweise die Baucbmuskula-
tur (welcbe ja nur teilweise dener^^ert war), in solchem Ausmass
geschadigt worden waren, dass sie auf dem Wege iiber die intakten
Kerven auf die peripberen Blutgefasse in der Leber einwirken
konnten.

Aus der obigen Bescbreibung der Yeranderungen bei durcb
Applikation des beissen Messingblecbs verursacbten Verbren-
nungen gebt bervor, dass die Muskulatur nur unwesentlicb ge-
schadigt worden war. Es ist recbt unwabrscbeinlicb, dass die
begrenzten Yeranderungen in der Muskulatur allein zu einer so
starken nervosen Eeizung fubren konnten, dass die Zunabme
der peripberen Blutmenge in der Leber dieselbe wiirde wie bei
denjenigen Yersucben, wo die Innervation innerbalb des gesamten
gescbadigteu Gewebsgebiets intakt war.

Um indessen unsere Scblussfolgerungen fernerbin zu sicbern
baben wir in einer Eeibe von Fallen mit dem obenerwabnten

310 GOSTA VON KEIS UND FRITIOP SJOSTRAND.

diinnen Kupferblech gebrannt, wobei pathologiscb-anatomiscb
nux die Haut, also ausscbliesslicb denerviertes Gewebe, gescha-
digt wurde. Diese Versucbe -mirden an 5 Tieren mit intakter
Hautinnervation nnd 4 Tieren, deren Haut und teilweise auch
Bauchmuskulatur denerviert worden war, ausgefiihrt. Die Ee-
sultate sind in Tab. 2 vriedergegeben.

Tnbelle 2.

Periphere Blutmenge in der Leber und Nierenrinde 45 Min. nach
Erzeugung einer besonders oberflachlichen Verbrennung.

Intakte Hautinnervation

Denervierte Haut

Nr.

Gewicht

Anzail Bintkorperchen
in tausenden pro
Gevrebe.

Nr.

Geivicht

Anzahl BlutkSrperchen
in tausenden pro mm^
Gewebe.

Leber

Nierenrinde

Leber

Nierenrinde

430

350

175

500

475

240

43*0

425

430

400

310

! 3

420

600

620

480

450

175

! 4

360

450

175

500

300

430

1 0

350

500

175

Die in Tab. 2 angegebenen Werte fUr die peripbere Blutmenge
in der Leber zeigen im Vergleich zu den entsprechenden Kontroll-
tierwert«n eine deutlicbe Zunahme, und diese Steigerung ist in
beiden Versucbsreiben von ungefahr derselben Grbssenordnung.

Es ist also vrahrscheinlich, dass eine derartige oberflacblicbe
Verbrennung von sovrobl intakter wie denervierter Haut zu einer
Zunabme der peripberen Blutmenge in der Leber fiibre, vresbalb
fiir wahrscheinlich gehalten werden kann, dass dieser Effekt obne
IMitwirkung der Innervation des geschadigten Gewebsgebiets
zustandekommen kann.

Ein Vergleicb der beiden Versuchsserien ergibt, dass der Effekt
bei Verbrennung denervierter Haut etwas Vt^eniger ausgesprocben
ist als bei intakter Haut, vas entweder darauf zuriickzufuhren
sein kann, dass die denervierten Tiere bei Schadigung von genau
gleichgrossen Hautbezirken "wie sonst grosser als die librigen
v'aren, oder moglicherweise ein Ausdruck dafiir sein kann, dass
ein Effekt auf dem V'ege fiber die Hautnerven neben dem auf
andere WeLse vermittelten Effekt vorliegt.

EINFI.USS LOKALER HAUTSGHADIGUNGEN.

311

Lokale Erfrierung von denervierter Haut wurde in 9 Fallen mil.
Chloratliyl und in 4 Fallen mittels geldihlten Jletalls hervorge-
Tufen. Die Jlittolwerte der Bestimmungen der periplieren Blut-
menge an diesen 13 Tieren warden fur die Leber 347,000 und fiir
die Nierenrinde 276,000 Blutkorperclien pro mm® Gewebe gegen-
iiber 510,000 bzw. 264,000 bei der entsprecbenden Bebandlung
von Haul mit xinversehrter Innervation. Diagramm 2 G demon-
strierfc die Streuung der Primarwerte.

Beim Yergleicli der Werte fiir die peripbere Blutmenge in der
Leber und Nierenrinde bei Kaltescbiidigung von denervierter und
intakter Haut wird ersichtlicli, dass die Zunabme in der Leber
im ersteren Fall nur 30 % gegeniiber fast 100 % im letzteren
ausmaclit. Die Verringerung der periplieren Blutmenge in der
Nierenrinde ist in beiden Fallen von derselben Grossenordnung.

Aus diesen Eesultaten geht hervor, dass der Effelct einer lokalen
Kiilteschadigung auf die peripbere Blutmenge in der Leber zum
grossten Teil von der intakten Hnutinnervation abhiingig ist und
somit durcb die Hautnerven vermittelt werden diirfte. Ein
gewisscr von der Hautinnervation unabbiingiger Effekt ist offen-
bar ausserdem wabrscbeinlich. Die eventuelle Verminderung der
peripberen Blutmenge in der Nierenrinde ist in beiden Fallen
<lieselbe, wcsbalb diese unabbangig von intakter Hautinnervation
zustandezukommen scbeint.

Erorterung dor Yersnclisergebnisse.

Friiher scbeint Simonart (1930) der einzige gewesen zu sein,
welcber das Verbalten der peripberen Blutgefiisse bei der Ver-
brennung direkt zu studieren versucbte, Seine nicbtquantitative
Metbodik gestattet indessen keine sicberen Scblussfolgerungen,
und zwar u. a. desbalb, weil Atbernarkose, in welcber er das Bren-
nen ausfubrte, an sicb erne allgemeine Dilatation dieser Blut-
gefasse, vor allem gerade in den Baucborganen, mitsicbbringt
(T. Sjostrand 1935, Lindgren 1935), wesbalb es scbuderig sein
muss, an Hand von einzelnen Beobacbtungen, um welcbe es sicb
bier bandelt, zu entscbeiden, ob die Dilatation eine Folge der Ver-
brennung oder beispielsweise einer verscbiedenen Narkosetiefe
bei dem gebrannten Versucbstier und dem Kontrolltier ist. Zudem
sagt eine derartige Metbodik nicbts iiber die Blutfiillung der Ka-
pillaren, sondern beriicksichtigt in erster Linie die Blutfullung in

312 • GOSTA VON REIS END FRITIOF SJOSTRAND.

den Arterioli und Ventdae, durch Tvelche die Farbe der Organe
hauptsachlicb bedingt wild.

ilan kann sicb vorstellen, dass die Variationen der peripberen
Blutmenge in der Leber (nnd Nierenrinde), welcbe bier bei Ver-
brennung und Erfrierung eines begrenzten Hautbezirks nacb-
gewiesen wurden, im Prinzip entweder durcb cbemiscbe oder
nervose Yermittlung, oder durcb einen direkten Temperaturein-
fluss auf das Parencbym der Organe zustandekommt.

Die Yersucbe mit Yerbrennung macben offenbar wabrscbeinlicb.
dass die Fernwirkung auf peripbere Blutgefasse in der Leber bei
dieser Form von Hautscbadigung unabbangig von der Innervation
des gescbadigten Gewebsgebiets eintritt. Da zudem ausgescblossen
ist, dass der Effekt die Folge einer direkten Erwarmung des Leber-
parencbyms ware, bleibt nur die Moglicbkeit iibrig, dass dieser
Effekt dutch chemische Vermittlung zustandegekommen ist.

Bei Yerbrennungen diirften somit innerbalb des gescbadigten
Gewebsgebiets aktive Substanzen gebbdet oder freigemacbt war-
den konnen, welcbe nacb ihrer Eesorption die peripberen Blut-
gefasse in der Leber direkt oder indirekt beeinflussen, sodass sicb
dieselben dilatieren, wabrend sie die entsprecbenden Blutgefasse
in der Nierenrinde entweder uberbaupt unberiibxt lassen oder
eventuell in geringerem Grade zur Kontraktion bringen. Die
Jluskelkapillaren, welcbe von diesen Substanzen nicbt beeintracb-
tigt warden, bffnen sicb in grosserem Ausmass durcb eine Ein-
■\virkung auf dem Wage iiber die Hautnerven.

Y'ir steben bier also bocbst komplizierten Yerbaltnissen gegen-
iiber, indem drei Organe auf ein und denselben Reiz, welcher den
Organismus trifft, ganz verscbieden reagieren. Es ist daber offen-
bar, dass Beobacbtungen uber Eeaktionen der peripberen Blut-
gefasse in einem gewissen Organ bei verscbiedenen Zustanden
keine generellen Scblussfolgerungen gestatten.

jMan kann fur wabrscbeinlicb balten, dass der bier nacbgewiesene
Effekt auf die peripberen Blutgefasse mit der Ausdebnung des
verbrannten Gewebsgebiets und der Dauer der scbadigenden
Einwirkung zunimmt, und dass die Dilatation dieser Blutgefasse
bei ausgedebnten Yerbrennungen einen solcben Grad erreicben
kann, dass dieselbe einen mehr oder weniger bedeutungsvollen
Faktor beim Zustandekommen des Yerbrennungsschocks dar-
stellen muss.

Dass es sicb dabei um einen primaren Faktor bandeln diirfte
scbeint daraus hervorzugeb’en, dass der Effekt auf die peripberen

EINFLUSS LOKALBR nADTSCIlADIQDNGEN. 3l;-i

Blutgefasse bercits bei einer Verbrcnming, vrelcbe keinen genii-
genden Umfang bat, um Scbock berbeizufiibren, und ausserdem
binnen so kiirzer Zeit, wie es bier der Fall ist, ein ganz aus-
gepragter ist.

Yor allem scbeint die Dilatation der peripberen Blutgefasse in der
Leber, wenn dieselbe bocbgradiger vsdrd, einen 'wicbtigen Scbock-
faktor bilden zu konncn. Ware diesc Annabme zutreffend, dann
batten "wdr wenigstens bier einen Scbockfaktor, welcber mit 'der
Intoxikationstbeorie im Einklang steben ^^'urde.

Der Effekt, wclcben man durch Resorption aus dem verbrann-
ten Hautgebiet erbalt, Aveist keine Abnlicbkeit mit demjenigen
auf, Avelcber durch eine Histamininjektion ausgelost wird, was
A'on Lixdgrex (1935) studiert worden ist. Die Annabme, dass
Histamin oder H-Substanz bei der Yerbrennung freigemaebt und
resorbiert wird, findet also an dicser Untersuebung keine Stiitze.

Die Result ate bei durch Erfrierung bervorgerufener ortlicher
Hautsebadigung Aveicben A'on denjenigen bei der Yerbrennung
insofern ab, als der Effekt auf die peripberen Blutgefasse in der
Leber zum grossten Teil durch die Hautnerv'en A’ermittelt AA'ird
und nur ein kleiner Teil auf andere Weise. Ein Effelct durch Ab-
kiiblung dcs Leberparencb}'ms ist ausgescblossen, da niebt ein-
mal ein relath* anbaltcnder Kaltereiz (sicbe unsere frubere Ar-
beit 1940) einen derartigen Effekt herbeifuhrt., obArohl die Ab-
kiihlung dabei noeb ausgesprochener sein muss. Derjenige Teil
der EinAAirkung auf die peripberen Blutgefasse in der Leber, Avel-
cher bei Erfrierung A'on denervierter Haut bestebenbleibt, diirfte
also durch cheniiscbe Yermittlung zustandekommen.

Die eA’entuelle Kontraktion A^on peripberen Blutgefassen in
der Eicrenrinde ist ja A'on einer intakten HautinnerA'ation unab-
biingig, und es bestehen somit Moglichkeiten eines ebemiseb
A’ermittelten Effekts auf diese Blutgefasse.

Yersuche mit lokaler Erfrierung und Yerbrennung baben er-
geben, dass der Effekt auf peripbere Blutgefasse in der Leber und
Nierenrinde in beiden Fallen gleicbartig ist; Avabrend aber die
Fernw’irkung bei der Yerbrennung hauptsacblich oder sogar A’^ollig
durch chemisebe Yermittlung zustandekommt, Avird dieselbe bei
Erfrierung bezuglicb der Leber zum grossten Teil v^on den Haut-
nerven und in geringerem jMasse ebemiseb vermittelt.

Denjenigen Effekten von Kaltescbiidigung, -welche Harkins
und Mitarbeiter (1934, 1935, 1937) nachgevdesen baben und die
denen bei der Yerbrennung ahnlich sind, konnen also aucb die

314 gSsta von BEIS UNJ) fbitiop sjSstrand.

Einwirkungen auf periphere Blutgefasse in der Leber und Nieren-
rinde angereiht werden, 'wobei bier gleicbzeitig auf den moglicben
UnterscMed der Entstebungsweise bingewiesen werden soil.

Die Versucbe mit lokaler mecbaniscber Hautverletzung zeigen,
dass eine rein mecbaniscbe Zerfetzung der Haut nicht zu einem
abnlicben Effekt wie Verbrennung oder Erfrierung fiibrt. Aus
diesem Eesultat diirfte zu entnebmen sein, dass ein blosser
Scbmerzreiz, welcber ja in diesen Fallen erbeblicb sein muss,
allein wabrend der Narkose keine betracbtlicbere Dauerwirkung
auf die peripberen Blutgefasse in der Leber und Nierenrinde aus-
uben muss.

Aus der Versucbsreihe, welcbe wir biermit bis auf weiteres
abgescblossen baben, scbeint bervorzugeben, dass die Hypotbese,
welcbe besagt, es werden durcb verscbiedene Hautreize (Laqueur
1930) und Scbadigungen (Lewis 1927) aus den Hautzellen Sub-
stanzen freigemacbt, welcbe nacb ibrer Resorption eine Allge-
meinwirkung auf z. B. die peripberen Blutgefasse des Organismus
ausuben konnten, nur binsicbtlicb der Verbrennung und mog-
licberweise der Erfrierung eine Stiitze findet. Dagegen lasst sicb
ein erbeblicber Einfluss auf die peripberen Blutgefasse in inneren
Organen bei mebreren verscbiedenartigen Hautreizungen durcb
Vermittlung der Hautnerven erzielen.

Die Ergebnisse diirfen zu der Annabme berecbtigen, dass die
cbemiscbe LabUitat, welcbe die Haut verscbiedenen dieselbe
treffenden Reizen gegeniiber aufweist .und die zu der LEWis’scben
H-Substanz-Tbeorie Anlass gegeben bat, lokal gebunden ist und
nicbt in grosserem Ausmass chemisch auf den iibrigen Organismus
einwirkt. Dagegen durfte diese cbemiscbe LabDitat dazu imstande
sein, die Vorbedingungen fiir die Reizung der Nervenendigungen
in der Haut bei z. B. Ultraviolettbestrablung zu scbaffen und auf
diese Weise auf dem Wege iiber die Hautnerven innere Organe zu
beeinflussen.

Die vorliegende Arbeit wurde durcb Unterstiitzung von Seiten
der Stiftung »Tberese ocb Joban Anderssons Minne« und des
Sonderfonds der mediziniscben Preisgruppe der Hobelstiftunc
ermoglicht.

EINFLUSS LOKALER HAUTSCnXDIQUNGEN.

315

Zusammcnfassung:.

1. Die Blutmenge in peripheren Blutgefassen der Leber und
Nierenrindc wurde 30 — 45 . Min. nach der Erzeugung entweder
einer oberflachliclien Yerbrennung, Erfrierung oder niecbanischen
Hautverletzung in cinem begrenzten Gebiet quer iiber der Bauch-
iiaut an Meerscbweinclien bestimmt. Bei den Versuchen rnit
Yerbrennung wurde nusserdem die Anzahl offener Kapillaren
pro Flacliencinlieifc des Quersclinitts im M. masseter bestimmt.
Die Yersuche sind in Pernoctonnarkose ausgefiihrt worden.

2. Bei lokaler Yerbrennung steigt die peripbere Blutmenge
in der Leber bis auf das im Jlittel reicblich zwcifaclie der ent-
spreebenden Kontrolltierwerte, wiibrend die entspreebende Blut-
menge in der Nicrenrinde etwas zuriickgebt oder iiberbaupt niebt
beeinflusst wird. Die Anzabl offener Kapillaren im Quersebnitt
des M. masseter nimmt um 25 % zu.

3. Bei derselben Bcbandlung von vorber denervierten Haut-
bezirken Avird gleicbwobl dicselbe Yerdopplung der peripberen
Blutmenge in der Leber crzielt, so^vie gleicbartige ^Yerte fur die
entspreebende Blutmenge in der Nierenrinde, aber keine Zunabme
der Anzabl offener Kapillaren im M. masseter.

Der Effekt auf die peripheren Blutgefiisse in der Leber scheint
durch Resorption einer oder einiger Substanzen aus dem geschii-
digten Gewebsgebiet vermittelt zu werden, welche direkt oder
indirekt dilatierend wirken, wabrend der Effekt auf die Kapillaren
im M. masseter auf dem ^Vege iiber die Hautnerven vermittelt
wird.

4. Bei lokaler Erfrierung steigt die Blutmenge in den peripberen
Blutgefassen der Leber auf das im Mittel doppelte der Kontroll-
tierwerte, wiibrend die entspreebende Blutmenge in der Nieren-
rinde um nahezu 30 % zuriickgebt.

5. Bei entsprechender Bcbandlung von denervierter Haut
erhiilt man eine Zunabme der peripberen Blutmenge in der Leber
von mix 30 %, wabrend die Yerminderung in der Nierenrinde
dieselbe ist wie bei den sub 4 erwabnten Yersueben.

Der Effekt auf peripbere Blutgefasse in der Leber scheint bei
ortlicber Erfrierung also hauptsiicblicb auf dem Wege iiber die
Hautnerven vermittelt zu werden, wabrend ein kleinerer Teil
der Wirkung auf peripbere Blutgefasse in der Nierenrinde che-
misch vermittelt werden diirfte.

22 — i01323, Ada phys. Scandinav. Vol. I.

316

GOSTA VON BBIS TJND FRITIOF SJOSTBAND.

6. Bei mechanischer Hautverletzung erhielt man als einzigen
Effekt ein Sinken der peripkereii Blutmenge in der Nierenrinde
um fast 40 %.

Schrifttum.

Dohrn, K., Dtsch. Z. Chir., 1901, 60 , 469.

Habkins, H. N., Proc. Soc. Exp. Biol., N. Y., 1934, 32 , 432.

— , Surg. Gynec. Obstet. 1938, 3 , 430.

— , und J. E. Hudson, Proc. Soc. Exp. Biol., N. Y., 1934, 32 , 434.
— , und P. H. Harmon, Ebenda, 1935, 32 , 1142.

— , — , J. Clin. Invest., 1937, 16 , 213.

ICrehl, L. und F. Marchand, Handbuch der allgemeinen Patbologie,
Bd. 1, Leipzig 1908.

Krogh, a., Anatomie und Pbysiologie der Capillaren, Berlin 1939.
Laquegb, a., Fortschr. d. Therap., 1930, 6 , 309.

Lewis, T., The blood vessels of the human skin and their responses,
London 1927.

Lindgren, a., Acta chir. Scand., 1935, 77 , Suppl. 39.

Pfeiffer, H., Allgeineine und experimentelle Pathologic, Wien 1927.
V. Eeis, G., B. P. Silfverskioed, F. Sjostrand und T. Sjostrand,
Skand. Arch. Physiol. 1938, 79 , 134.

V. Reis, G. und F. Sjostrand, Ebenda, 1937, 77 , 71.

— , — , Ebenda, 1938, 79 , 139.

— , — , Acta physiol, scand., 1940, 1 , 183.

SiMONART, A., Arch. int. Pharmacodyn., 1930, 37 , 269.

Sjostrand, T., Skand. Arch. Physiol, 1934, 68 , 160.

— , Ebenda, 1935, 71 , Suppl. 5.

SONNENBURG, E., Dtsch. Z. Chir., 1878, 9 , 138.

Stockis, E., Arch, int. PharmacodjTi. 1903, 11 , 201.

Wilms, M., Mitt. Grenzgeb. Med. Chir., 1901, 8 , 395.

Wilson, W. C., Medical Research Council Report Nr. 141, London 1929.

From the Physiologj^ Institute, Helsingfors University, and
the Laboratory of Neurophysiology, the Caroline
Institute, Stockholm.

Electrotliyrcograin, Blood Iodine and Thyroid
Iodine during Stimulation of
the Sympathetic.^

B. HELIN and H. ZILLIAOUS.

In 1914 Cannon and his collaborators started a series of in-
vestigations, among vliich is wortli mentioning elicitation of
th 3 Trotoxical symtoms bj’ means of continuous faradic stimula-
tion of the sympathetic in the neck or bj* effecting an anasto-
mosis between the sympathetic and the phrenic nerve. In 1916
Cannon and Cattell showed that faradic stimulation of the
sympathetic or an intravenous injection of adrenaline caused an
action potential in the thyroid. Bueget (1917) and Maeine
and his collaborators (1918) repeated the experiments of Can-
non and his collaborators, but could not produce thyrotoxic
sj-^mptoms by means of stimulating the sympathetic. Haney
(1932), however, was able to establish a definitely increased
metabolism by these means, a result, the correctness of which
is contested by Feiedgood and Bevin (1939).

Several workers have made sections of all the nerves leading
to the thyroid and then noted changes such as alterations in the
normal chloride content and the normal indigo-carmine excre-
tion. In recent years Nakaoka (1934), Peczenik (1935) and
others have made histological observations after section or sti-
mulation of the sympathetic. In the former case they noted
atrophic glands, in the latter hyperplastic ones. Rossine, Keit-
CHEVSKAiA and Semenov (1936) recorded action potentials from
the thyroid after stimulation of the sympathetic and after in-

* Received 15 October 1940.

318

B. HELIN AND H. ZILLTACOS.

section of adrenaline. They emphasize the dependence of the
jhyroid secretion on both nervous and humoral factors. Ha-
tAiiA ( 1936 ) also recognizes in the action potentials, which he
calls clectrothyreograms, evidence of the sympathetic participa-
ting in the regulation of the secretory activity of the thyroid.

The Problem.

It has been proved anatomically as well as histologically and
to some extent physiologically that many nerves, mostly of a
sympathetic, but also of a parasympathetic nature, lead to the
thyroid. Numerous tests have shown that these nerves affect
the vasomotor processes in the gland, and many experiments tend
to prove that the nerves, especially the 83rmpathetic nerves, in-
fluence glandular secretion, but some tests have jdelded nega-
tive results. A study of these facts has induced us to under-
take an investigation into the correlation between action po-
tentials (or electrothyreograms), blood iodine and thyroid iodine.
Our problem is therefore: does stimulation leading to an action
potential of the thyroid, also cause blood iodine and the iodine
•content of the thyroid to deviate from the normal in such a fash-
ion that the electrothyreogram can be considered as an index
of secretory activity in the gland? The second problem, a natural
consequence of the first question, was: is the correlation between
the electrothyreogram, blood iodine and thyroid iodine of such
a nature as to provide a further argument in favour of the
sympathetic innervation of the thyroid gland?

All electrothyreograms hitherto recorded have been elicited
by sympathetic stimulation, either direct by faradic stimulation
or by means of an intravenous injection of adrenaline. In no
case have action potentials been noted after stimulating the
■v'agus or its branches. In view of these circumstances we 'have
used sympathetic stimulation in all our tests, either direct
faradic excitation of the nerve or an intravenous injection of
adrenaline.

The questions raised divide the experiments into the follow-
ing sections:

registration of electrothyreograms during

a) faradic stimulation of the sympathetic in the neck

b) an intravenous injection of adrenaline

ELECTROTH YREOGRAM .

319

measurements of iodine content in

a) tlie blood after faradic stimulation,

b) tbe blood after an intravenous injection of adrenaline

c) normal blood,

d) normal tli)T:oid gland substance,

e) stimulated tbyroid glands.

Electropliysiological Technique.

The action potentials were led off from decerebrate cats^ according
to descriptions by Cakkox and Cattell and by Hasama.

The animals were operated in ether anaesthesia. All contact bet-
ween tracheal cannula on the one hand, the tbyroid, blood vessels
and nerv'es in its ^*icmity on the other, is carefully avoided. By re-
leasing the carotid compression during decerebration so much blood
is allowed to collect in the evacuated skull that 10 cem can be with-
drawn with a syringe for the iodine analysis to be made later in the
experiment, whereupon the cavity is dried and filled with cotton
wool in the usual manner. After the operation the cat is left lying
for two hours. Silvcrchloride electrodes with cotton contacts are used.
The leads arc taken to a balanced resistance-coupled push-pull amp-
lifier connected to a string galvanometer. One electrode rests on the
surface of the thyroid, the other is indifferent, e. g., in the decerebra-
tion wound. Electronegativity in the gland corresponds to upward
movement on the film. The central stump of the sympathetic is sti-
mulated immediately below the superior cervical ganglion.

The method of recording electrothyreograms, when stimulatmg
with adrenaline is the same as above except that now adrenaline (1
cem 1: 70 — 100,000) is injected into one of the femoral veins.

Results.

a. The Eleotrotbyroogram during Faradio Stimulation
of the Sympathetic.

After compensating the demarcation current until a constant
baseline is attained, tbe sympathetic is stimulated faradically
on one side for ^/, — 1 minute (distance between tbe coils 7 — 11
cm.). Tbe gland regularly becomes electronegative in relation
to tbe neutral point. A slow, continuous deflection .of the string
can be seen during which the potential difference attains its

' To avoid obtaining values in the subsequent iodine analyses that might be
due to food rich in iodine, the test animals were placed on regular diet poor in
iodine for five days before the experiments.

320

B. HELIK AKD H. ZILLIACUS.

maximum, about 1 mV, in 2 — 3 minutes, thereafter to decrease
continuously until the baseline is reached. The whole of this
potential change lasts 5 — 8 minutes. With a few exceptions the
gland, after stimulation, proved to be electro-negative. Fre-
quently the curve did not return to the baseline, but remained
slightly above or below the original one (Fig. 1).

Pig. 1. Fractions of the electrothyreogram during faradic stimulation of the
sympathetic. Calibration with 1 mV. Black vertical lines mark film stop.

Fig. 2. The electrothyreogram during intravenous injection of adrenaline.
Calibration with 1 mV.

b. The eleetrothyTeogram during Intravenous Injection

of Adrenaline.

Adrenaline that is administered intravenously (1 ccm 1: 90,000)
regularly causes electro-negativity in the gland. A latent period
of 5 — 10 seconds precedes a slow and continously increasing
potential difference which attains its maximum in about 2 mi-
nutes. The descending phase of the curve sets in just as after
faradic stimulation, then proceeds continuously, and stops slight-
ly above or below the original baseline. (Fig. 2).

The thyroid potentials after sympathetic stimulation, de-
scribed above, correspond very well to those recorded by Can-
non and Catteli. and by Hasama. These authors, by means of
convincing tests, came to the conclusion that the vasomotor
process is not the cause of the action potentials and consider
that the slow and continuous potential difference is related to
the process of secretion in the gland, a statement, to which we
did not wish to subscribe without further argument. A parallel

ELBCTROTHYREOQRAM. 321

increase in blood iodine and a reduced iodine content in tbe
thyroid gland may be regarded as such arguments.

Tecliniquo of Iodine Determination.

As already stated, tbe iodine determination in tbe blood and
tbyroid gland form tbe second part of tbe experiment. For tbis
purpose Yre employed a modified Leipert method of micro-iodine
analysis wbich vra bad subjected to special investigation. (He-
Lix, ZiLLiAcus and Unomus, 1939).

a. Determination of the Iodine Content of Blood after
Faradie Stimulation of the Sympathetic,

As it vras a question of determining tbe iodine content in tbe
blood of tbe test animal under the same conditions as during
the recording of tbe electrotbyreogram, faradie stimulation or
adrenaline injections ■were carried out in exactly tbe same man-
ner as before. In some cases "we were even able to use tbe same
test animals for both phases of tbe experiment. Tbe general
procedure has been described above; during decerebration exactly
10 cem of blood are •witbdra'wn •with a syringe and after a couple
of hours tbe sympathetic •was stim'ulated (coil distance 7 — 10 cm)
faradically immediately belcw tbe superior cer-vical ganglion
three to five times at intervals of 40 minutes. In about half tbe
cases tbe electrotbyreograms were photographed. After tbe last
period of stimulation blood samples were taken at once for two
or more analyses from one of the carotids, after which tbe experi-
ment was concluded by an analysis of these samples. In order
to avoid disturbing prevalent conditions of equilibrium in tbe
tissues we did not take blood samples between tbe periods of
stimulation. (Table 1).

b. Determination, of the Iodine Content in Blood after an
Intravenous Injection of Adrenaline.

These tests were made under tbe same experimental condi-
tions as before. The adrenaline (1 cem 1; 70 — 100,000) was in-
jected intravenously into tbe femoral vein and tbe interval be-
tween tbe injections was 40 minutes. During decerebration 10
cem of blood were taken for determining normal iodine, after

322 B. HELIX AXD H. ZILLIACDS.

Table 1.

Blood iodine content after faradic stimulation of the sympathetic.

Blood

iodine

(y n)

before

stimnl.

Blood iodine %)

Faradic stimulation

after stimulation

periods

Increase

Test animal.

in blood

Cat No.

Indi-

vidual

Average

Num-

ber

Dura-

tion,

Strength

iodine
in %

values

min.

23.4

27.5

30.1

28.8

23.1

33.7

33.0

33.1

33.7

2.1

34.3

27.6

24.8

28.4

28.3

14.1

29.1

■

14.2

7.4

15.4

14.9

101.3

15.8

16.1

3.7

17.8 .

16.7

351.4

16.9

14.8

8.1

13.6

66.6

12.4

21.1

31.6

28.6

30.0

8 •

42.1

Aver.

17.3

23-7

3.8

1.1

36.6

wliicli the cat was allowed to lie for two hours before the first
injection was made. After three to five injections accompanied
by registration of gland potentials the experiment was concluded
by an iodine analysis of two or more blood samples, all of which
were therefore taken after the last injection had been made
(Table 2).

A glance at the two tables at once shows a distinct difference
between the normal blood iodine values on the one hand and
the values obtained after faradic stimulation of the sympathetic
or after intravenous injection of adrenaline on the other. The
averages already display an increase in the blood iodine after
stimulation, but, as the normal blood iodine content of the dif-
ferent animals varies considerably (3.7 — 30.0 y %), the values for
each individual animal are of greater importance. In no case
it was possible to record a reduction of the blood iodine content.

ELECTROTHYREOQRAM.

323

Table 2.

Blood iodine content after intravenous injection of adrenalin.

Test animal

Blood

iodine

before

inject.

Blood iodine (y *)
after injection

Adrenalin injections

Increase
in blood
iodine ^

Cat No.

Indi-

vidual

values

Average

Num-

ber

Quan-

tity

ccm

Concentra-

tion

16.6

21.8

20.4

21.x

1:80000

27.8

6.9

19.2

19.4

19.8

1 ; 100000

179.6

14.0

22.5

20.7

21.6

1:100000

54.2

10.2

14.6

13.3

13.6

1:100000

33.3

16.7

19.4

21.2

18.8

19.8

1:100000

18.6

11.3

10.6

9.8

10.6

1:80000

41.9

11.9

17.7

15.9

16.8

1:70000

41.1

Aver.

11.9

17.6

3.8

1:90000

47.1

0. Determination of the Iodine Content of Normal Blood.

The loss of blood during the operation and withdrawal of the
first blood sample for the control causes a reduction of the blood
volume, and it might therefore be assumed that the increased
value obtained after stimulation was due to a mobilization of
blood from a blood reservoir in the thyroid or to other changes
in the internal milieu. Decerebration is also a process that might
well be supposed to affect the blood iodine values. We there-
fore carried out a number of controls, in which, in accordance
with the earlier experiments, the iodine content of the blood
was determined at the time of decerebration, after which the
animal was allowed to lie approximately the same time that
was occupied in performing a complete stimulation test after
which the iodine content of the blood was determined afresh. As
the table indicates, exceedingly small variations occur in the
normal blood iodine, so that it must be held certain that the

324

B. HELIN AND H. ZILLIACUS.

increased iodine values are due to sympathetic stimulation and
not to some other circumstance in connexion with the perfor-
mance of the tests. (Table 3).

Table 8.

The Mood iodine content at the time of decerebration and 4 — 5 hours
later, no stimulation of the thyroid having occurred
in the interval.

Iodine con-

Iodine content (/ %) of the

(

Test animal

Cat No. •

tent (/ %)
of the blood

blood 4 — 5 hours later

Blood iodine
increase or

at time of
decerebration

Individual

values

Average

decrease in %

19.8

■ 1

19.G

20.4

19.3

19.9

+ 1.5

19.9

; 2

10.5

10.3

9.7

lO.O

— 5

15.9

! 3

15.9

16.2

16.1

+ 1

16.8

22.1

' 4

22.0

22.7

22.4

— 0.9

22.4

13.9

i ^

14.4

15.7

14.8

— 4.1

14.8

•

Aver.

16.8

16.6

- 1.2

d. Determination of the Iodine Content of Normal and
Stimulated Thyroid Glands.

As a further check on our results we extirpated both the thy-
roid lobes of each test animal and determined their iodine con-
tent separately. As the iodine content of the same gland cannot
be determined both before and after stimulation and as the
iodine content of a normal thyroid varies considerably in the
case of each animal and even in the different lobes of the same
gland, only a comparison' between the average iodine content
of a large number of stimulated and unstimulated glands pos-
sesses any significance. As both modes of stimulation, that we
used, take effect through the sympathetic .and as blood iodine

BLECTROTU YREO G R A SI .

325

in both cases is increased in mnch the same manner and degree,
Ave have not, in compiling table 4, differentiated between faradi-
zation and injection of adrenaline. In order to obtain values for
unstimulatcd thyroids, we extirpated the thjToid glands of a
dozen cats and analysed their iodine content (Table 4).

Table 4,

Iodine content in normal and stjmpatheticalhj siimvlatcd thyroid gland

substance.

Imliiie content

Iodine content in

in normal

thyroids after sym-

thyroids

pathetic stimnlation

(?i J)

0.00140

0.00190

O. 002 ir

0.004S0

O.oioon

O.OlOlO

0.03370

0.00970

0.020G1

0.01481

0.0157t

O.oifisn

0.01G80

0.03370

0.0550S

0.020GI

0.02437

O.OOlOf)

0.0127.3

0.00142

0.00204

0.01021

0.00318

0.00726

O.00S13

O.OIS.GI

O.0U12

0.0108G

0.071-12

0.00482

0.01510

0.00-193

Average: 0.01319

0.00521

0.00998

0.0143G

0.01878

O.OOIGC

O.0072S

Average: O.OlOSS

It is interesting in this connexion to mention that Rahe, Ro-
gers, Fawcett and Beebee (1914) established a difference in
the iodine content between the stimulated and unstimulated
lobes of the same thyroid as early as in 1915.

According to our results, shown in table 4, the normal iodine
content of a cat’s thjToid gland varies a great deal and averages
0.01919 % (fresh thyroid gland substance) and after sympathetic
stimulation 0.01035 %. The variations indicate that this proce-
dure does not lend itself very well to checking our conclusions.

326

B. HELIN AND H. ZILLIACUS.

Discussion.

A survey of the experiments indicates that increase in the
iodine content of the blood — by 36.9 % in case of electric sti-
mulation and by 47.1 % in case of stimulation with adrenaline
— means that the thyroid gland product has been transferred,
to the blood. And as it was possible to record an action potential
from the thyroid under the same conditions of stimulation that
caused the altered iodine values referred to, it must be held
that the electro thyreogram and the increased blood iodine are
different aspects of the same active process in the gland. This
correlation scarcely leaves any room for doubt as to the electro-
thyreogram after sympathetic stimulation being an index of sec-
retory activity in the thyroid. The experiments at the same
time prelude a further argument in favour of the thyroid gland
secretion being dependent, at any rate partly, on the sympa-
thetic nervous system.

Sununary.

Electrothyreograms elicited in decerebrate cats by faradic
stimulation of the sympathetic or by injection of adrenaline have
been recorded with the aid of a string galvanometer and a direct-
ly-coupled balanced push-pull amplifier. The aim of this work
has been to correlate blood iodine, thyroid iodine and electro-
thyreograms, and, if possible, to obtain evidence for sympathetic
control of the thyroid gland.

Blood iodine has been determined by a modified Leipert me-
thod, described in detail in a previous communication (Helin,
ZiLLiAcus, Unonius, 1939).

Blood iodine was found to be increased after stimulation of
the sympathetic as well as after injection of adrenaline. In the
former case the increase averaged 36.9 %, in the latter case
47.1 %. Controls with unstimulated animals showed that de-
cerebration and operation around the thyroid did not as such
influence the iodine concentration of the blood.

The increase in blood iodine after stimulation was found to
be accompanied by a decrease of thyroid iodine amounting to
about 85.4 %.

ELECTROTH YBEOGRAM .

327

Stimulation bringing about .these changes in the iodine con-
centration of the gland and the blood is always accompanied by
nn electrothyreogram which accordingly is regarded as an index
of secretory activity in the gland.

Beferences.

Subget, G., Amer. J. Physiol. 1917, 44 , 492,

Cannox, V., C. A. L. Singer and R. Fm, Ibidem 1914, 36 , 363.
•Cannon, V., and M. Cattell, Ibidem 1916, 41 , 39.

■Cannon, V. and R. Fitz, Ibidem 1916, 40 , 126.

■Cannon, V. and D. Rapport, Ibidem 1921, 68 , 308.

Cannon, V. and P. Smith, Ibidem 1922, 60 , 466.

Feiedgood, H. and S. Bevin, Amer. J. Physiol. 1939, 125 , 153.
Hasama, B., Pfiug. Arch. ges. Physiol. 1936, 238 , 758.

BDcnce, F. and A. Haney, Amer. J. Physiol. 1932, 102 , 249.
Helin, B., H. Zileiacus and E. Unonios, Nord. Med. 1939, 4 , 3580.
Marine, D., M. Rogoff and 6. Stewart, Amer. J. Physiol. 1918,
45 , 268.

Hakaoha, I., Ber. ges. Physiol. 1934, 78 , 115.

Peczenik, 0., Pfiug. Arch. ges. Physiol. 1935, 235 , 486.

Rahe, j., j. Rogers, G. Fawcett and S. Beebee, Amer. J. Physiol.
1914, 34 , 72.

Rossine, j. a. Kritchevskaia and N. Semenov, Ibidem 1936, 97 ,
631.

Aus der pharmakologischen Abteilung des Karolinischen Instituts

Stockholm.

tiber die Wirkiing you Digitalis, Cardiazol,
Coramiiij Hexeton und Strychnin auf Kreislaiif
und Atiniing des gesunden Mensclien.^

Von

G. LILJESTRAND nnd G. NYLIN.

Die therapeutisch gebrauchten Elreislaufmittel uben ihre Wir-
kung bekanntlicb in. sehr verscbiedener Weise aus. Mancbe von
ibnen beeinflussen im Wesentlichen nur das Herz selbst, indent
sie es zu kraftigeren Kontraktionen veranlassen. Diese Gruppe
wird vor allem. von den Digitalissubstanzen vertreten; nacb der
gelaufigen Anscbauung scbeinen diese Stoffe nur dann eine
Verbesserung des allgemeinen Kreislaufs zu bewirken, wenn das
Herz insuffizient und somit ausserstande ist, das in den grossen
Venen zuruckstromende Blut in normalem Umfange weiterzu-
befordern. Andere Mattel dagegen greifen extracardial an, indeni
sie entweder durcb periphere Wirkung auf die Gefasse oder
durch Stimulierung des Vasomotorenzentrums den Gefasstonus
steigern. Sie konnen dadurch unter Umstanden das venose
Angebot an Blut vermebren und infolgedessen das Minutenvolu-
men des Herzens steigern. Als Beispiele fiir Kreislaufmittel, die
peripber, an den Gefassen selbst angreifen, sind vor allem Adre-
nalin, Epbedrin und verscbiedene Oxyepbedrine zu erwabnen.
Allerdings iiben diese Substanzen bisweilen gleicbzeitig aucb eine
unmittelbare Herzwirkung aus. Ausserdem bedingen sie eine
Erbobung des Stoffwecbsels, wodurcb ebenfalls indirekt der
ICreislaufapparat zu vermehrter Arbeit angeregt wird. Die
Zunahme des Minutenvolumens ist bierbei jedocb relativ grosser
als die Stoffavecbselsteigerung, so dass man sie nur teilweise als

* Der Redaktion am 1.0. Okfober 1940 zngegatigen.

KRBISLAUFMITTEL TJKD HERZMINDTENVOLUMEN. 329

eine Folge der Stoffwechselwirkung anseken darf. Endlich, konnea
als typisclie Jlittel mit direkter Einwirkung auf die nervosen
Zentren Cardiazol, Hexeton und Coramin, so\rie Strychnin ange-
fiihrt -werden. Sowohl Atem- wie Vasomotorenzentrum werden
hierbei primar gereizt; das Herz wird aber nur indirekt, z. B.
dutch Verbesserung der Blutstromung beeinfiusst (vgl. Hilde-
BRAXDT, 1937).

Fiir viele der erwahnten Mittel, die bei insuffizientem Krelslauf
ausgedehnte VerT\’endung finden, sind keine oder nur sparliche
quantitative Angaben beziiglich ihrer "Wirkung am Menschen
bekannt. Es sind zvrar zahlreiche Tierversuche mitgeteilt worden,
die eine giinstige Kreislaufvirkung demonstrieren, es muss aber
daran erinnert verden, dass in diesen Fallen so gut wie immer
^’iel hohere relative Dosen gebraucht vrurden, als sie bei der
therapeutischen Behandlung kranker Menschen in Frage kommen.
Ausserdem sind die Bedingungen oft vereinfacht, indem regula-
torische Einrichtungen teilweise ausser Spiel gesetzt -wurden.
Es ist somit von praktischem Interesse, direkte Bestimmungen
am Menschen auszufuhren. Um eine feste Basis fiix Untersuchun-
gen an Patienten mit insuffizientem Kreislauf zu schaffen, sind
Versuche an gesunden Jlenschen wiinschenswert, in der Hoffnung
hierdurch zu einem besseren Verstandnis des Wirkungsmechanis-
mus zu gelangen. Fiir Adrenalin und verwandte Substanzen
wurden solche Bestimmungen schon friiher ausgefuhrt (v. Euler
und Liljestrand 1927 und 1929, Liljestrand und Linde 1933,
Berggren und Soderberg 1938). In der vorliegenden Mitteilung
soli uber entsprechende Untersuchungen mit Digitalis sowie mit
den zentral angreifenden Substanzen Cardiazol, Coramin, Hexeton
und Strychnin berichtet werden.

Unsere Versuche wurden an drei Studenten im Alter von
20 — ^25 Jahren ausgefuhrt. Es wurden immer Standardbedingun-
gen eingehalten. Die Versuchspersonen kamen in niichternem
Zustande ins Laboratorium, sassen in vorsatzlicher Bluskelruhe
etwa eine Stunde, ehe mit den Bestimmungen begonnen wurde
und blieben selbstverstandlich wahrend der ganzen Versuchszeit
im Zustande vollkommener Ruhe. Mundstiick und Nasenklemme
%vurden wahrend des ganzen Versuches nicht abgenommen. Das
Jlundstiick war an einem Dreiwegehahn befestigt und konnte
dutch Drehen des Hahnes entweder mit einem Lovenventil
oder mit einem Gummibeutel (Fussballblase) verbunden werden.
Dutch das Exspirationsventil wurde die Luft zu einem grossen

330

G. LILJESXBAND DND G. NYLIN.

Tabelle

Digiioial 1 ml intravp.nds, entsprechend

Vor der Injektion

1.6 ml.

241.8 ±
8.0

64.0 +

1.6 I

15—30 Min.

Oj-Ver-
branch
ml pro
Min.

Pnlsfre-
qnenz
pro Min.

238

233

258

221

254

264

Tabelle
Oardiazol 1 ml

Versuchs

person

Pnlsfre-
quenz
pro Min.

_60

59.9 +

1.6

Vor der Injektion

Eednz.

15—30 Min.

tnde X
Pnla-
freqnenz

Minnten-
volnmen
des Her-
zens 1.

Oj-Ver-
branch
ml pro
. Min.

Pnlsfre-
qnenz
pro Min.

3.4

234

4.2

254

4.8

246

59 i

4.6

226

60 1

4.8

274

3.9

251

3.8

250

3.6

248

4.7

264

4.8

268

4.14 ±

251.6 ±

58.8 +

5.16

4.7

1.7 1

KREISLAUPMITTEL END HERZMINUTBNVOLEMEN.

331

1 .

0.1 g Folia Digitalis.

nacb fler Injektion

40 — 60 ilin. nach der Injektion |

j Blnt-

1 druck

Rednz,
ImpE-
tnde X
Pals-
frequenz

Minnten-
volnmen
des Her-
ZCDS 1.

Oj-Ver-
brancb
ml pro
Min.

Pnlsfre-
qnenz
pro Min.

Blut-

dmck

Rednz.
Ampli-
tude X
Pnls-
freqnenz

Minnten- j
volnmen i
des Her- !
zens 1. j

: 116/77

24.3

4.6

236

122/82

25.6

3.6 :

' 119/86

20.2

4.5

237

119/86

21.2

3.6

1 97/62

29.1

3.8

270

99/62

31.3

5.6

—

3.8

222

—

3.2

1 _

—

4.2

242

—

3.9 i

j 121/87

19.3

3.9

264

121/88

18.6

4.4

4.05 +

245.2 ±

63.0 +

4.03 ±

0.20

7.4

1.9

0.34

10 °/o suhlcuian.

nacb der Injektion

—60 Min. nacb der Injektion

i Blut-
j drnck

Rednz.
Ampli-
tude X
Pnls-
freqnenz

Minuten-
volnmen
des Her-
zens 1.

Oj Ver-
braucb
ml pro
Min.

Palsfre-
qncnz
pro Min.

Blut-

drnck

Rednz.
Ampli-
tude X
Puls-
frequenz

28.2

4.2

232

122/85

26.4

3.9

i 114/76

24.8

3.9

241

119/77

27.8

4.0 1

24.0

3.7

239

100/68

22.0

3.1

25.7

3.8

102/66

25.7

4.5

( ~

—

4.7

276

—

4,0

! —

—

4.4

246

—

3.8 i

? —

—

4.0

256

—

4.3

—

4.1

238

—

4.6

—

4.4

264

—

4.7

—

5.0

273

—

4.9

4.22 +

249.0 ±

59.5 ±

4.18 ±

0.13

5.6

1.8

0.17

23 — i01323. Acta phys. Scandinav. Vol.I.

332 G . ULJJESTRAND TJND G . NYLIN.

Tabelle
Coramin 1 ml

Vor der Tnjektion

— 30 Min.

Versochs-

person

Oj-Ter-
brauch
ml pro
Min.

Pulafre-
quenz
pro Min.

Blnt-

drnck

Eednz.
Ampli-
tude X
Pnls-
freqnenz

Minuten-
volnmen
des Her-
zens 1.

02 -Ver-
braucb
ml pro
Min.

Pulsfre-
qnenz
pro Min.

A. E. . .

222

117/79

26.4

3.9

230

i . .

227

. 66

119/86

21.2

3.3

238

D. T. . .

218

22.6

3.7

218

> . .

212

—

4.8

195

> . .

232

—

3.8

246

> • •

219

—

3.8

221

H. L. . .

270

—

4.6

299

> . .

258

—

3.9

253

232.S +

65.1 ±

3.98 ±

237.6 ±

64.6 +

7.3

1.3

0.17

10.9

1.4

Spirometer geleitet. Der Gaswechsel ■wnrde durch Sammlung der
Ausatmungsliift ■walirend 6 — 10 Minnten und nachheriger Gas-
analyse im Haldaneapparat bestimmt. Zur Ermittlung des Herz-
minutenvolumens -WTirde dxirch, den erwahnten Habn die Ver-
bindnng mit dem Gummibeutel bexgestellt, der vorber mit einem
passenden Gemiscb aus Acetylen, Sauerstojff und Stickstoff ge-
fiillt worden war. Die Bestimmung gescbab nacb der Metbode
von Gbollman (1932). Die Pulsfrequenz wurde durcb Palpation
der Arteria radialis, der Blutdruck durcb Auskultation, unter Ver-
wendung des von Liljestrand und Zander (1928) bescbriebenen
Manometers festgestellt. Jlit dem Produkt aus Pulsfrequenz und
der sogenannten reduzierten Amplitude, d. b. des Pulsdruckes,
dividiert durcb den mittleren arteriellen Blutdruck, erbielten wir
ein relatives Mass fiir die Zirkulationsgrosse (vgl. Liljestrand
und Zander 1928, Aperia 1940).

Jeder Versucbstag wurde mit Normalbestimmungen eingeleitet.
Dann wurde die zu priifende Substanz injiziert, und weitere Be-
stimmungen erfolgten innerbalb der nacbsten 15 — 30, bzw. 40 —
50 blinuten.

Als geeeignetes Digitalispraparat benutzten wir Digitotal, das

KREELAUFMITTEL UND HERZMDJUTENVOLDMEN.

333

25 /• sttbhuian.

nach der Injektion

^40 — 60 Slin. nach. der Injektion

Blut-

dnicfc

Eedaz.
Ampli-
tude X
Pols-
frequenz

Minuten-
■volumen
des Her-
zens 1.

Oj-Ver-
brauch
ml pro
Hin.

Pnlsfre-
qnenz
pro Min.

Blnt-

drnck

Rednz.
Ampli-
tude X
Pnls-
frequenz

Minnten-
volnmen
des Her-
zens 1.

120/79

27.2

4.9

232

121/77

28.8

3.9

IW/W

19.9

4.0

233

123/85

23.7

3.5

23.0

3.4

218

100/70

21.6

3.4

—

3.6

204

—

3.0

—

5.1

218

—

4.4

3.7

218

—

3.6

—

4.8

285

—

5.3

—

3.9

248

—

3.9

4.18 ±

232.4 ±

64.3 +

3.88 ±

0.2S

9.2

1.3

0.26 1

intravenos gegeben "vnirde. Von diesem Praparat entspricbt 1 ml
0.1 g Digitalisblattern; die in tinseren Versucben venvendete
Dosis vrar 1, bz'w. (in einem Versncb) 1.5 ml. Die Brgebnisse
sind in Tabelle 1 zusammengestellt.

Die Pulsfrequenz zeigt nacb 15 — 30 iMinnten eine statistiscb
nicbt geniigend sicbergestellte Abnabme nnd der Blntdruck,
sowohl nacb 15 — 30, als aucb besonders nacb 40 — 60 IMinnten
eine nnbedeutende Erbobung, die gewobnbcb so^vobl den systo-
liscben als aucb den diastoliscben Druck betrifft. Das Ampli-
tudenfrequenzpxodukt bleibt aber praktiscb genommen unver-
andert. In tibereinstimmung damit ist das Minutenvolumen des
Herzens innerbalb der Peblergrenzen konstant. Dasselbe gilt
fiir den Sauerstoffverbraucb.

Die von Stetvart (1931) mit der Metbode von Grollman
beobacbtete, scbver zu verstebende Abnabme des IMinutenvolu-
mens bei Gesunden nacb Digitalisznfubr, baben mr also nicbt
bestatigen konnen. Allerdings bat Stewart diesen Effekt erst
spat (viele Stunden) nacb peroraler Vexabreicbung gefunden; es
ist aber zu erwarten, dass die Wirkung nacb intravenoser Zufubr
besonders scbnell und deutlicb eintritt. Mit der Metbode von

334

G. LILJESIRAND OND G. NYLIN:

Tabelle
Hexeion 2 ml 10 %

Yersuclis-

person

Vor der Injfektion

—30 Min.

Oj-Ver-
branch
ml pro
Min.

Blut-

drnck

Eednz.
Ampli-
tude X
Pnls-
freqnenz

Minuten-
volnmen
des Her-
zens 1.

Pnls-
frequenz
pro Min.

A. K. . .

224

113/82

19.1

3.6

231

> . .

222

108/79

18.8

3.7

246

D. T. . .

218

—

3.9

222

H. L. . .

240

—

4.6

270

> . .

266

—

4.3

276

234.0 ±

59.8 ±

4.00 +

249.0 +

59.2 +

8.6

1.7

0.17

1.6

Tabelle

Strychninnitrat

Vor der Injektion

-30 Min

Versnchs-

person

Oj-Ver-
branch
ml pro
Min.

Puls-
frequenz
pro Min.

Blut-

druck

Reduz.
Ampli-
tude X
Puls-
frequenz

Oj-Ver-
branch
ml pro
Min.

Pnls-
frequenz
pro Min.

H. L.^ . .

260

125/92

18.3

4.9

256

> - . .

248

119/89

16.4

4.1

245

. 3 . .

. 228

116/86

20.0

3.7

223

245.3

58.3

4.23

241.3

57.3

Broemser hat Hartl (1932) an einem Kjeislaiifgesnnden nach
intravenoser Zufuhr von 0.0005 g Strophantin eine erhebliche
Abnahme des ]\Enutenvolumens gesehen, gegen die Methode
konnen jedoch ernste Bedenken angefiihrt werden (vgl. Aperia
1940).

^ 1 mg Strychninnitrat subkntan.

® 1.6 mg Strychninnitrat snbkutan.
® 2 mg Strychninnitrat snbkutan.

KREISLAUFJitlTTEIi END HERZMINUTENVOLUMEN.

intramushuldr.

335

nBch der Injektion

—60 Min. nack der Injektion

Blnt-

drnck

Reduz.
Ampli-
tude X
Pnls-
frcquenz

Minnten-
volnmen
des Her-
zens 1.

Oj-Ver-
hrauch
ml pro
Min.

Pnls-
freqnenz
pro Min.

Blnt-

drncfc

Rednz.
Ampli-
tude X
Pnls-
freqnenz

Minnten-
volnmen
des Her-
zens 1.

112/81

19.8

3.6

250

114/82

20.9

3.4

117/89

16.4

4.S

230

117/81

23.2

4.1

—

3.7

236

—

3.8

—

4.4

244

—

3.9

—

4.9

238

—

4.6

4.1G±

239.6+

61.6 +

3.84 +

0.26

3.4

2.2

0.26

0 .

subHatian.

nack der Injektion

40—60 Min. nack der Injektion

Blnt-

drnck

Rednz.
Ampli-
tude X
Pnls-
freqnenz

Minuten-
volnmen
des Her-
zens 1.

0,-Ter-
branck
ml pro
Min.

Pnls-
freqnenz
pro Min.

Blnt-

drnek

Rednz.
Ampli-
tude X
Pnls-
frequenz

Minnten-
volumen
des Her-
zens 1.

128/90

20.1

5.0

274

129/90

20.8

5.6

122/87

19.4

3.8

243

124/87

20.2

3.8

118/87

16.9

3.9

232

121/89

17.6

3.9

4.23

249.7

58.0

4.40

Cardiazol (Pentamenthylentetrazol) mirde in einer Menge von
1 ml einer 10 % Losimg subkutan injiziert; nur in einem Falle,
dem letzten der Tabelle 2, wurde eine Dosis von 1.4 ml gebraucht.

Wie aus Tabelle 2 bervorgeht, weisen weder Sauerstoffver-
braucb und Herzminutenvolumen, nocb Pulsfrequenz oder Blut-
druck irgend eine sichere Veranderung auf. Das Ergebnis stebt
somit im Gegensatz zu den Resultaten von Hoen und Neuthard
(1937), die — ebenfalls nacb 1 ml der 10 proz. Losung subkutan

336 Q. LILJESTRAND DND G. NYLIN.

— eine betraclitliclie Erhohung des Minutenvolumens beobach-
teten. 20 Minuten nacb der Injektion fanden sie eine Zunabme
des Stromvolumens um. 48 %. Diese Zunabme war teilweise
durcb Erhobung des Stoffwecbsels (um 17 %), teilweise aber
durcb scblecbtere Ausniitzung des Blutsauerstoffes bedingt.
Leider wurden aber ibre Versucbe nicbt unter Standardbedin-
gimgen ausgefiibrt, was die Beweiskraft der Resultate erbeblicb
einscbrankt. Auffallend ist aucb, dass 10, bzw. 30 Minuten nacb
der Injektion keine Zunabme, sondern sogar nacb 10 Minuten
eine Abnabme des Minutenvolumens von 11 % konstatiert wurde.
Diese Tatsacbe stebt in scblecbter Ubereinstimmung mit den
Ergebnissen anderer Versucbe, aus denen man scbliessen kann,
dass die Resorption des Cardiazols nacb subkutaner Injektion
rascb erfolgt.

Coramin (Pyridin-/5-carbonsaurediatbylamid) wurde regelmassig
in Mengen von 1 ml der 25 proz. Losung eingespritzt. Wie Tabelle
2 zeigt, sind bier ebenfalls Stoffwecbsel und Herzminutenvolumen
sowie Pulsfrequenz und Blutdruck konstant geblieben. Eiir
Coramin liegen ebenfalls Bestimmungen von Hoen und Neu-
THARD vor. In ibren Versucben war 20 Minuten nacb der Injektion
von 1 ml der 25 proz. Losung der Sauerstoffverbraucb um 16 %
und das Herzminutenvolumen um 127 % erbobt. Aucb diese
Untersucbungen wurden, ebenfalls wie die mit Cardiazol, nicbt
unter Einbaltung der Standardbedingungen ausgefiibrt. Sie
zeigten regelmassig in den ersten 20 Minuten eine Steigerung des
Stromvolumens, die dann aber wieder zuriickging. Sebr bemer-
kenswert ist, dass die Pulsfrequenz etwas verlangsamt wurde, so
dass das Scblagvolumen, das vor der Injektion 68 ml betrug,
auf der Hobe der Wirkung den ausserordentlicb boben und
unwabrscbeinlicben Wert von 169 ml erreicbte. In unseren Ver-
sucben findet man keine Andeutung einer derartigen Wirkung.

Entsprecbende Versucbe an gesunden Menscben mit Hexeton
(Metbylisopropylzyklobexenon) scbeinen bis jetzt nicbt publiziert
worden zu sein. Aus Tabelle 4 ist zu erseben, dass 2 ml der 10
proz. Losung, intramuskular injiziert, obne Wirkung bleiben.
Die Verbaltnisse liegen also genau wie beim Coramin.

Die Wirkung von 1 — 2 mg Strycbninnitrat wurde an einer Ver-
sucbsperson, bei subkutaner Injektion gepriift (Tabelle 5). Aucb
bier ist das Ergebnis vollkommen negativ. Unter Benutzung der
bereits erwabnten Broemser’scben Metbode fand Hartl (1932)
bei einem Patienten mit Cancer oesophagi eine Zunambe des

KMEISI-AUFMmEL UND HERZMINUTENVOLOMEN. 387

Minutenvolumens um 37 %, 15—30 Mnuten nack der subkutanen
Applikation yon i mg Strychnin, Wir konnen diese Beobachtung
nicht bestatigen.

Es wurde schon ftuher hervorgehoben, dass die stimulierende
Wirkung des Cardiazols und des Goramins auf das Atemzentrmn
bekannt ist. Man konnte dies nicht nur in Tierversuchen, sondern
anch beim Menschen, nach e^erimentell herabgesetzter Atmung
beobachten. An gesunden Versnchspersonen, bei denen die
Eunktion des Atemzentrums durch eine vorherige Gabe von
0.02 g Morphin vresentlich herabgesetzt vrorden "war, haben
Stedtinger nnd Gaubatz (1935), sovrie Stanton Hicks (1935)
nach subkutaner Znfuhr von 1 ml der 10 proz. Oardiazollosung,
bzvr. der 25 proz. Coraminlosung eine bedeutende Steigerimg der
Atmung gesehen. Da vrir in unseren Versuchen die Exspirations-
luft analysierten, also den respiratorischen Quotienten feststellten,
vrar es uns moglich zu entscheiden, ob eine Zunahme der Atmung
unter den von uns gewahlten Bedingungen entstand. Trifft dies
zu, so ist zu erv’arten, dass eine gewisse Ausvraschung von Kohlen-
saure stattfindet, ivodurch der respiratorische Quotient erhoht
vrerden muss. Aus Tabelle 6 geht hervor, dass keines der ge-
priiften Mittel eine sichere Anderung des respiratorischen Quo-
tienten bevrirkt hat.

Tabolle 0.

Respiralorischer Quotient vor und nach Injehtion verschiedener Stoffe.

S t 0 f f

Zahl

der

"Ver-

snche

Respiratorischer Quotient

ror der
Injektion

15 — 30 Min.
nach der
Injektion

40—60 min.
nach der
Injektion

Cardiazol

0.801 ± O.007

0.788 ± O.012

0.770 ± 0.0I8

Coramin

0.806 ± O.018

0.810 + 0.0S6

0.771 + 0.022

Hexeton

0.782 ± 0.016

0.794 ± 0.028

0.788 ± 0.026

Strychninnitrat . , .

0.807

0.827

0.788

Digitotal

0.798 + 0.024

0.798 ± 0.011

0.797 ± 0.016

IJnsere Versuche haben eindeutig ergeben, dass bei Verabrei-
chung in den gebrauchlichen therapeutischen Dosen am gesunden
Menschen, keine der von uns untersuchten Substanzen eine Bin-
irirkung auf Kreislauf oder Atmung ausiibt, Nach unserer Ansicht
steht diese Tatsache in bester tfbereinstimmung mit den Vor-

338

G. LIUESTRAND UND G, NYLIN.

stellungen, die man sich xiber die Wirkungsweise dieser Stoffe
gebildet bat. Am gesunden Menscben ist die Eegulation der in
Frage stebenden Funktionen so gut eingestellt, dass den ge-
wobnlicb gebraucbten, verbaltnismassig kleinen Dosen keine
Wirkung zukommt. Damit soil natiirlicb nicbt gesagt sein, dass
die Verwendung derartiger kleiner Dosen bei Zustanden mit
insuffizientem Kreislauf nutzlos sei. Ibre Wirkung in diesen
Fallen muss abet besonders gepriift warden.

Znsammenfassimg.

Therapeutiscbe Dosen von Digitalis (Digitotal), Cardiazol,
Coramin, Hexeton und Strychnin wurden gesunden Versucbs-
personen unter Standardbedingungen injiziert und zeigten keine
Oder nur sebr scbwacbe Wirkung auf Pulsfrequenz, Blutdruck,
Herzminutenvolumenj Sauerstoffverbraucb und respiratorischen
Quotienten,

Iiiteratup.

Aperia, a.: Skand. Arch. Physiol. 1940, 83 . Suppl. 16.

Berggren, S., und L. SonERBERo; Ebenda 1938, 79 . 116,

V, Euler, U. S,, und G, Liljestrand: Ebenda 1927. 52 . 243.

— , — , Ebenda 1929, 55 . 1 .

Grollman, a.: The cardiac output of man in health and disease,
London 1932,

Hartl, K.: Klin. Wschr, 1932. 11 . -356,

Hildebrandt, F.: Handb. exp, Pharmakol. Erg-werk, 1937, 5 .
128 und 151.

Hoen, E,, und A. Neuthard: Arch, exp. Path. Pharmak. 1937. 185 .
302.

Liloestrand, G., und P, Linde: Arch, int. Pharmacodyn. 1933, 45 .
318.

Liljestrand, G,, und E. Zander: Z. ges. exp. Med. 1928. 59 . .105.
Stanton Hicks, C.: Austral. J. exp. biol. a, med. Sci. 1935. 13 . .^61.
Stbininger, H., imd E, Gaubatz: Klin. Wschr. 1935, 14 . 169,
Stewart, H, J,: Proc. Soc. exp. Biol., N. Y., 1931, 29 . 209.

Aus der pharmakologischen Abteilung des Karolinischen Institutes

Stockholm. ’

Bie Wirkung des Atropins aiif das HeTzminiiteii-
Tolnmen des gesnnden Menschen.^

Ton

R. DOMENJOZ.

Infolge der Schliisselstellung seiner Angriffspunkte ist die EiiX'
wirkung des Atropins auf den Kreislaufapparat ausserordentlich
komplexer Natur. Neben der Beeinflussung der Herzfrequenz
und der Art des Kontraktionsablaufes, sind aucb die Effekte
dieser Substanz auf die Kreislaufperipherie, z. B. auf die Kapillaren
(v. Lederer, 1936) fiir die Gesamtwirkung bedeutungsvoll. Bei
der haufigen und vielfachen Anwendung des Atropins und seiner
galeniscben Bereitungsformen, ganz besonders aucb bei Affektio-
nen des Zirkulationssystems, iiberrascbt es einigermassen, dass
liber seine Gesamtwirkung, "wie sie in integralbafter Weise im
Herzminutenvolumen zum Ausdruck gelangt, keine einbeitlicben
Vorstellungen herrscben.

Die Veranderungen der Eorderleistung des Herzens unter dem
Einfluss von Atropin sind schon offers Gegenstand experimen-
teller Untersucbungen gewesen. Zum ersten mal wtirde eine Stei-
gerung des !Minutenvolumens von Odaira (1926) beobacbtet. In
Versucben an Kanincben mit der EiCK’scben Methode konnte
dieser Autor nacb intravenoser Verabreicbung von Atropinsulfat
in Mengen von 0.2 bis 0.55 mg pro kg Tier, eine Zunabme der Herz-
leistung zwiscben 4: und 33 % konstatieren. Diese Eeststellungen
konnten jedocb von anderen Untersucbem niobt bestatigt warden.
MarshaIiL (1926), der ebenfaUs mit der FicK’scben Methode die
Korrelationen zwischen Herzfrequenz und Schlagvolumen stu-
dierte, konnte bei Gaben von 3 — I mg keine Zunabme des hlinuten-

‘ Eingegangen am 1. November 1940.

R. BOMENJOZ.

340

volumens registrieren. Marshall verwendete zu seinen Versuclien
trainierte Hunde, die eine Durclifuhrung der Untersucliungen
ohne Narkose ermoglickten. Auch Harrison, Blalock, Pilcher
und Wilson (1927) sahen bei Steigerungen der Pulsfrequenz zwi-
scben 100 und 400 % des Ausgangswertes nur Anderungen in
der Forderleistung des Herzens, die sick innerhalb der Febler-
grenzen der verwendeten Metbode bewegten, Im Mittel fanden
sie, sowobl an morpbinisierten, als auch an unnarkotisierten
Hunden eine Zunabme des Minuten volumens von 4 %.

Prinzipiell ist zu diesen Tierversucben zu sagen, dass die
Genauigkeit der damit erreicbten Resultate nur gering ist, sodass
unbetracbtlicbe Steigerungen des Minutenvolumens wegen der
weiten Fehlergrenzen statistiscb nicht einwandfrei nachgewiesen
werden konnen. Der Grund bierfiir liegt wobl darin, dass die
Standardbedingungen, zumal bei unnarkotisierten Tieren, nur sebr
unvollkommen eingehalten werden konnen. Weiterhin ist die
Bedeutung der Reaktionslage fiir die Wirkung vegetativer Gifte
von so ausscblaggebender Bedeutung, dass eine Ubertragung der
an Tieren gevronnenen Resultate auf die Verbaltnisse am Menscben
nur mit Vorbebalt moglicb ist.

Die Frage der Einwirkung des Atropins auf das Herzminuten-
volumen des Menscben erfubr eine erstmalige Bearbeitung durcb
SraTH, Burwell und be Vito (1928). Das bierbei verwendete
Material bestand aus gesunden Studenten, an denen das Minuten-
volumen nacb der Metbode von Field und Bock bestimmt
wurde. Atropin wurde intravenos, in Mengen von 1.2 mg ver-
abreicbt. Die in diesen Versucben beobacbteten Steigerungen der
Pulsfrequenz geben bis zu 40 % imd die Zunabme des Minuten-
volumens bis zu 10 % der Ausgangswerte.

Metbodik.

Unsere Versucbe "wurden an 2 gesunden Brwaobsenen, in niicb-
ternem Zustande, bei vollkommener, -willkurlicber Muskelrube
durcbgefiibrt. Zu .Beginn jeder Untersuobung wurde eine Rube-
periode von ungefabr 1 Stunde eingescbaltet, wabrend der die
Versucbsperson gut bedeckt in einem Liegestubl sass. Darauf
wurde mit Bestimmungen der Rubewerte in folgender Reiben-
folge begonnen; Pulsfrequenz, Blutdruck, Sauerstoffverbraucb und
Herzminutenvolumen. In jedem Versucb wurden 2 Normal-
bestimmungen durcbgefiibrt. Atropinsulfat wurde in einer Menge

ATROPIN €ND HERZMINDTENVOLDMEN.

341

von 0.001 g siibkutan injiziert und die Wirloing an Hand der
Pulsfrequenz verfolgt. Die -weiteren Bestimmungen gescbaben in
wechselnden Intervallen, 30 bis 90 Minuten nacb der Injektion.

Die Versucbsperson atmete mittels eines Kautscbulnnnnd-
stiickes durcb einen vreiten Dreiwegebabn, der die Verbindung mit
einem Spirometer oder mit einem Gummibeutel ermoglichte, der
das Acetylengemiscb entliielt.

Die Blutdruckmessung erfolgte nacb der auskultatoriscken
Metbode mit Hilfe des von Liljestrand und Zander (1928) be-
scbriebenen Metbylenjodidmanometers.

Zur Ermittelung des Sauerstoffverbraucbes wurde die Exspira-
tionsluft vabrend 7 bis 8 Minuten in einem grossen Spirometer
gesammelt und die Ventilationsgrosse pro Zeiteinbeit bestimmt.

Die Bestimmung des Minntenvolumens erfolgte nacb der Ace-
tylenmetbode von Grollman (1932). Das von uns verwendete
Gasgemiscb batte folgende Zusammensetzung: 7 L Luft, 3 L
Narcylen und 0.5 L Sauerstoff. Die Entnabme der Proben gescbab
nacb 5 — 6, bzv. nacb 8 — 9, tiefen, rascb aiifeinanderfolgenden
Atemziigen, entsprecbend einer Zeitspanne von 18 — 20, bzw. von
25 — 30 Sekunden.

Neben dem Minutenvolumen wurde zur Abscbatzung der Herz-
leistung nocb das sogen. »reduzierte Amplitudenfrequenzprodukt«
nacb Liljestrand und Zander (1928) bestimmt. Diese Grosse,
in den Tabellen als A. F. P. vermerkt, ergibt sicb als Produkt von
Pulsfrequenz und sogen. reduzierter Amplitude, vrobei unter
reduzierter Amplitude das Verhaltnis des Pulsdruckes (Amplitude)
zum Mitteldruck (aritbmetiscbes Mittel aus systoliscbem und
diastolischem Druck) zu versteben ist.

Eesultate.

Die in unseren Versucben gefundene Steigerung desHerzminuten-
volumens erscbeint gering, venn man sie mit Werten vergleicbt,
wie sie z. B. bei Verabreicbung von Adrenalin und seiner Ver-
wandten (von Euler und Liljestrand (1927 und 1929), LiljE'
STRAND und Linde (1933)) oder gar in Aibeitsversucben zutage
treten. Fiir die Bevreisfaaft derartiger Untersucbungen ergibt
sicb bieraus, dass das peinlicbe Einbalten der Standardbedingungen
wabrend der ganzen Versucbsperiode von ausscblaggebender
Bedeutung ist. Wir baben diesem Faktor ganz besondere Bedeu-
tung gescbenkt. In jedem Versuch wurden, nacb einer ausgiebigen

342

E. DOMEEJOZ.

Tabelle

Verdnclerungen von Blutdruch, Pulsfrequenz, Sauerstoffverbrauch,

O.ool g Atropinsulfai.

Nor

m a 1

■w e r

t e

Datum

Blut-

druck

Pnls-

frequ.

Mitt el
dmck

A.F. P.

Saner-

stoff-

verbr.

Art.-

venose

Diff.

Min.-

Vol.

15. X.

115/82

98.5

18.4

224

58.1

3.9

115/83

17.8

224

57.4

3.9

17. X.

119/88

103.5

16.8

216

52.8

4.1

123/94

108.5

15.0

218

52.7

4.1

19. X.

119/89

104

16.5

208

52.3

4.0

119/89

104

16.6

215

52.7

4.1

Mittelwerte:

217.6

4.0

Streunng :

± 6.1

± 0.1

Tabelle

Verdnderungen von BlutdrucJc, Pulsfrequenz, Sauerstoffverbrauch,

0,001 g Atropinsulfai.

Nor

m a 1

w e r

t e

Datum

Blut-

druck

Puls-

frequ.

Mittel-

druck

A.P. P.

Sauer-

stoff-

verbr.

Art.-

venBse

Diff.

Min.-

Yol.

21. X.

109/76

92.6

20.7

211

57.1

3.7

120/88

104

16.0

212

59.8

3.6

23. X.

105/75

19.8

205

56.8

3.6

114/84

17.6

205

59.6

3.4

25. X.

110/78

21.1

207

55.0

3,8

120/82

101

21.8

197

57.4

3.4

Mittelwerte;

206.0

3.6

, Streunng:

± 5.6

±0.17

1 .

ATROPIN UND HERZMINUTENVOLTJMEN.

343

Utilisation und Eerzminntenvolumen nach suhkutaner Injehtion von
{Versuchsperson R. D.).

1 ~ ^

j 'VVorto nach Injektion von O.OOl Atropinanlfat

Zeit nach
Inj.

Blut-

drnck

Ptils-

freqn.

Mittel-

drnck

A. F. P.

Saner-

stoff-

verbr.

Art.-

venOse

Diff.

Min.-

Vol.

35'

118/86

20.1

227

51.8

4.4

60'

122/92

23.0

225

4.5

47'

123/89

106

21.6

224

43.2

5.2

64'

123/98

110.6

17.2

224

52'

124/95

109.5

17.7

217

42.9

5.1

71'

124'94

109

20.4

218

44.0

5.0

Miltelwert der Verandernng des Minntenvolnmens:

= + 21.2 «

i Znnahmo der Pnlafreqnenz zvischen 16 nnd 50

Utilisation und Eerzminutenvolumen nach subhutaner Injehtion von
iVersuchsperson L. G.).

UVerte nach Injektion von O.OOl Altropinsulfat

Zeit nach
Inj.

Blut-

druck

PuIb-

freqn.

Mitttl-

druck

A. P, P.

Sauer-

stoff-

verbr.

Art.-

venOse

Diff.

59'

126/88

107

29.1

220

49.1

4.6

72’

129/94

111.6

25.1

215

48.1

4.6

72'

123/88

105.6

24.8

207

51.6

4.0

102'

129/94

111.6

22.6

209

55.4

3.8

59'

117/86

101.6

24.4

204

38.4

5.8

81'

126/92

109

24.8

206

50.4

4.1

Jlittelwert der Terandernng dee Minntenvolumens:

= + 21.8 >6

Zunahme der Pulsfrequenz zwischen 24 und 48

346

R. DOMENJOZ.

Literatur.

V. Euler, U. S., und G. Liljestrand: Skand. Arch. Physiol. 1927.
52. 243.

— , Ebenda 1929. 55. l._

Grollman, a.: The cardiac output of man in health and disease,
London 1932.

Grosse-Brockhofp, F., imd F. Kaldenberg: Arch. exp. Path. Phar-
mak. 1938. 188 . 383.

Harrison, T. E., A. Blalock, C. Pilcher und Ch. P. Wilson: Amer.
J. Physiol. 1927. 83 . 284.

V. Lederer, E.: Arch. exp. Path, Pharmak. 1936. 182 . 362. 372.
Liljestrand, G., und E. Zander': Z. ges. Med. 1928, 59 . 105.
Liljestrand, G., und P. Linde: Arch, int, Pharmacodyn. 1933. 45 . 318.
Marshall, E. K.: Amer. J. Physiol. 1926 a. 77 . 459,

— , J. Pharmacol. 1926 b. 39 . 167.

Odaira, T.: Tohoku J. exp. Med. 1925, 6 . 325.

Plavsic, C.: Arch. mal. coeur. 1939. 32 . 163.

Smith, W. C., S. Burwell und M. J. de Vito: J. din. Invest. 1928. 6 .
237.

From the Institute of Theoretical Physics, University of Copenhagen.

Rate of Penetration of Ions through
the Capillary WaU.^

L. HAHN and Gr. HEVESY.

("With 3 figures in the text.)

In this paper, the results of experiments arc communicated
which were carried out in order to get information on the rate of
passage of the ions of important constituents of the plasma as
sodium, potassium, chlorine, and phosphate through the capillary-
wall. Crystalline substances introduced into the circulation will
soon invade the extracellular fluid of the body. On this fact
is based the method usually applied to determine the size of the
extracellular space. Sucrose, sulphocyanate, or sulphate in-
troduced into the human circulation were found (Lavietes et
alia 193G), for example, to be completely distributed between the
plasma and the tissue space in the course of two or three hours.
A complete distribution of thiocyanate in the extracellular space
of rabbits in the course of half an hour is recorded (Keogh 1937).

The partition of a substance introduced into the circulation
between plasma and the extracellular fluid involves two processes:
(1) penetration across the capillary wall and (2) distribution by
diffusion and convective processes in the capillary and the extra-
cellular fluids. The last mentioned processes will play a secondary
role, only, in -view of the very short distances between the capil-
laries. Taking the length of the distances involved (Keogh
1926) to be less than 60 /{ and the diffusion coefficient of the sub-
stance investigated to be at least 1 cm® per day, the time necessary

^ Received 11 November 1940.

24 — i01323. Acta phys. Scandinav. Vol. I-

348 I* HAHN AND G. HEVESY.

to displace, for example, a sodium ion from one end of the ca-
pillary space to the other, or from one end of the corresponding
extracellular space to the other, will be less than 2 seed We arrive
at this result by considering the propagation by diffusion only
of the substance which penetrated the capillary -wall. The fluid
is, however, not without a circulation of its own, and this circula-
tion will possibly shorten the time arrived at in the above cal-
culation.

By introducing some sodium chloride into the circulation and
by measuring the time it takes for a certain fraction to leave the
circulation it should be possible to measure the rate of passage of
sodium chloride through the endothelium. However, when carry-
ing out these experiments we meet the folio-wing difficulties:
(a) Not only does the circulation get rid of excess sodium chloride
by giving off salt to the extracellular space, but also by taking
water up from the tissues. Keys (1937) found, when studying
the fate of sucrose intravenously injected into man, that osmotic
equilibrium by a shift of -n ater takes place from three to ten times
as fast as sucrose exchange. The rate of disappearance of the excess
sodium chloride -will, thus, not measure the rate of passage of
sodium chloride through the capillary wall but a more complex
process, (b) We do not measure by the method outlined the rate
of passage of sodium through the endothelium but a resultant of
the rate of passage of sodium and chloride. The resistance of the
endothelium to the passage of sodium and chloride may be quite
different, (c) The introduction of appreciable amounts of sodium
chloride into the circulation will disturb the normal conditions
prevailing in the circulation. When one tries to eliminate this
difficulty by introducing small amounts only, the analjiiical
difficulties become almost unsurmountable. All these difficulties
can be eliminated by injecting into the veins labelled sodium
chloride (sodium chloride containing some radioactive ®*Na of
negligible weight) and by measuring the rate of disappearance of
the active ions from the plasma, i. e. the decrease of the radio-
acti-vity of the plasma. We are not determining in these experi-
ments the rate of influx of excess sodium chloride from the plasma
into the extracellular fluid but the rate of exchange between
labelled plasma sodium and non-labelled extracellular sodium, as
the number of sodium (®^Na ®*Na) atoms of the plasma remains

^ The mean displacement of a particle r = V2 D, where D is the diffnsion
coefficient.

HATE OF PENETRATION OF IONS.

349

practically constant all tlirough tlie experiment. The rate of
exchange -will be determined by the permeability of the capillary
wall to sodium ions and will, thus, be a measure of this perme-
ability.

We carried out also experiments with radio-potassium, radio-
chlorine, radio-bromine, and radio-phosphate, while heavy water
was used as an indicator for water in the study of permeability
of the endothelium to water. The measurement of the distribution
of radio-sodium between plasma and body of the rabbit was
pre^dously used to determine the extracellular volume of the
rabbit (Griffith and Margraith 1939; Hahn et alia 1939).

Experimental Procedure.

Badioactive sodium and potassium were prepared by bombard-
ing NaOH and KOH, respectively, with high speed (10 million
volts) dcuterons. The hydroxydes were neutralized with hydro-
chloric acid and the solution thus obtained was injected. Kadio-
active chlorine and bromine were prepared by bombarding NaCl
and NaBr, respectively, with deuterons. The active chlorine and
bromine obtained verc driven off as HCl and HBr, respectively,
and were collected in a sodium hydroxyde solution. This proce-
dure was chosen to get rid of the active sodium simultaneously
produced with the active halogens. We are much indebted to
Dr. J. C. Jacobsen and hir. 0. N. Lassen for preparing the
radioactive substances by making use of the Copenhagen
cyclotron.

About 3 cc. solution containing the radioactive substances of
an activity of about 1 microCurie was applied. The salt concentra-
tion of these solutions was brought up to a physiological level by
adding non-active sodium chloride. The solution was injected
into the jugularis of the rabbit and blood samples of about 1 cc.
were collected at intervals from the carotis. Plasma samples
of known weight were dried and their radioactivity was compared
by using a Geiger counter, Bor comparison of the radiocativity of
plasma and muscle samples the samples were ashed at about
400° and the plasma ash mixed with non-active muscle ash of the
same weight as the corresponding active muscle ash sample.
Blood and muscle samples were secured simultaneously from the
narcotized rabbit.

350 I-. HAHN AND 6. HBVESY.

Fig. I. Rate of disappearance of various labelled ions from the plasma.

Eesults.

The results obtained are seen in Tables 1 to 5 and Figs. I to III.
The tables contain data on the percentage of the labelled element
injected still present in 1 cc. plasma at various intervals. The
volume of diluting fluid necessary to bring down the concentra-
tion of the substance injected to that found after a given time is
also stated. Furthermore, this diluting volume is expressed in
percent of the rabbit’s body weight.

We shall first compare the rate of disappearance- of sodium,
chlorine, and bromine from the circulation. This comparison
encounters no difficulties since practically the sole outlet^ of these
elements from the circulation is the extracellular body fluid,
though some **Na is taken up by the surface layer of the bone

. * In experiments taking up to 1 hour, the amount of *^Na lost by excretion
is much less than 1 per cent of the amount administered.

BATE OF PENETRATION OP IONS.

351

Min.

Fig. n. Rate of disappearance of Tarious labelled ions from the plasma.

apatite (Hahn et alia 1939). No great difference is found between
the rate of passage of sodium, chlorine, and bromine through the
capillary wall and the values obtained for different rabbits show
fairly large variations. These variations are to some extent due
to differences in the size of the extracellular space which is found
by numerous experimenters to show quite appreciable differences
for different rabbits.

A comparison of the rate of passage of potassium, phosphate,
and water with that of sodium, chlorine, and bromine encounters
some difficulties since potassium, and the same applies to phosph-
ate and water, has an additional outlet into the tissue cells in
contrast to the first mentioned group. The amount of *-K lost
by the^blood after the lapse of a given time is the resultant of
the amount penetrated into the tissue fluids and that returned
from the latter into the blood. When besides the interspaces the
cellular space opens an outlet to the <=K leaving the circulation,
the amount returning from the tissue fluids into the blood will
be reduced and, thus, the resultant concentration of the plasma
will be lowered. Though the potassium content of the cells is

352

L. HAHN AND 0. HBVESr.

Table 1.

Bate of disappearance of ^^01 from the circulation of rabbits
weighing 8.3 and 8.4 kg, respectively.

Time in min.

Percent of ^*01 in-
jected present in

1 gm. plasma

Diluting fluid volume

in cc. (appa-
rent extracel-
lular volume)

in percent of
body -weight

-•

Rabbit I.

0.37

0.622

161

6.4

0.73

0.486

206

8.2

1.01

0.476

211

8.6

1.48

0.408

245

9.8

2.06

0.400

250

lo.o

3.8

0.329

304

12.2

10.6

0.224

446

17.8

18.6

0.188

532

21.3

0.182

550

22.0

Rabbit II.

0.8

0.62

161

6.7

0.9

0.28

■ 357

14.9

2.6

0.174

575

24.0

4.7

0.165

607

25.8

8.8

0.161

622

26.0

16.6

0.160

668

27.8

26 ■

0.143

700

29,2

0.128

783

32.6

only partly replaced by *^K during the experiment (Hahn et alia
1939) in view of the low potassium content of the plasma and the
high content of the tissue cells the additional outlet opened' by
the intrusion of into the cells in experiments taking one hour
makes out about five times the normal outlet of intrusion of these
ions into the interspaces.

The total water content of the cells can be entirely replaced by
labelled water. Since the volixme of the cellular body water is
about twice as large as that of the extracellular fluid through
the intrusion of labelled water into the cells a substantial
additional outlet of the labelled water molecules- of the plasma
is opened.

RATE OP PENETRATION OP IONS.

353

TaWe 2.

Bate of the disappearance of from the circulation of a rabbit

weighing S.7 kg.

Time in min.

Percent of
injected

present in 1 gm.
plasma

Diluting fluid volume

in cc.

(apparent ex-
traceUnlar vo-
lume)

in percent of
body veight

1.0

0.37

270

lO.O

2.2

0.27

370

13.7

8.1

0.21

475

17.6

16.S

0.18

556

20.6

0.14

715

26.6

58.5

0.12

835

30.9

Table 3.

Bate of disappearance of ~^Na from the circulation of rabbits
weighing 2.7 kg, and 2.4 kg., respectively.

Time in min.

Percent of '*Na in-
jected present in

1 gm. plasma

Diluting fluid volume

in cc. (appa-
rent extraceb
Inlar volume)

in percent of
body vreigbt

Bab it I.

0.2

0.80

125

4.6

0.45

0.5 0

199

7.4

0.9

0.41

242

9.0

1.5

0.32

310

11.0

2.2

0.30

331

12.3

3.8

0.234

427

15.8

5.2

0.215

466

17.3

0.194

515

19.1

Babbit II.

1.8

0.63

158

6.6

2.3

0.282

357

14.9

, 4.3

0.218

458

19.1

9.1

0.181

553

23.0

0.163

615

25.6

0.161

663

27.6

0.140

715

29.8

120

0.134

745

31.1

354

L. HAHN AND G. HEVEST.

Table 4.

Bate of disappearance of from the circulation of rabbits
weighing B.5, 2.4 and 2.8 Teg, respectively.

Percent of injec-

ted present in 1 gm.
plasma

Diluting fluid volume

Time in min.

in cc. (appa-
rent extracel-
lular volume)

in percent of
body weight

Babbit I.

O.G

0.19

526

21 1

2.0

0.082

1220

49 1

0.068

1470

59 i

0.068

1720

69 {

0.036

2860

114

Babbit II.

0.088

1730

10.6

0.048

2080

20.6

0.038

2640

110 1

40.6

0.031

3220

134 I

0.0277

3610

151 1

210

0.0269

3720

155 1

Babbit III

{

0.26

0.364

283

12.3 i

0.60

0.220

455

19.8 i

1.06

0.129

775

33.7

3.06

0.067

1755

76.3

As to the phosphate ions, not onlj that they diffuse into the
tissue cells but they are also incorporated into the surface layer
of the bone apatite. These additional outlets may be made
responsible for the fact that the loss of the plasma is found to
be very much greater than the ®*Na loss during the same time.

While the additional outlets mentioned above for potassium,
phosphate, and water will be responsible for the high values of
the volume of the diluting body fluid observed for these elements
in experiments of comparatively long duration, the fact that also
after the lapse of ^4. minute only, larger amounts of potassium
than of sodium, chlorine, or bromine are lost by the plasma
requires another explanation. After such a short time the volume

355

RATE OF PElfETRATIOIT OF IONS.

Table 5.

Bate of disappearance of ^~P from the circulation of rabbits
weighing 2.1 and 2.7 hg., respectively.

Time in min.

Percent of *-P injec-
ted present in 1 gm.
plasma

Diluting fluid T-olnme j

in cc. {appa-
rent extracd-
Inlar rolnme)

in percent of
body weight

Babbit I.

O.soo

333

15.9

i 1.S

0.234

427

20.4

i 3.0

0.187

535

25.5

4.5

0.14S

699

33.3

' 6.8

0.112

892

42.5

10.9

0.081

1230

58.6

; 16.9

0.060

1670

79.5

1 25.9

0.04G

2180

104

1 39.0

0.0S2

3130

149

Babbit IT.

i 0.2

0.63

160

5.9 1

; 0.45

0.46

216

8.0 ;
1

! 0.9

0.S5

287

10.6 '

! 1.5

0.26J

383

14.2 1

I 2.2

0.234

428

15.9 1

i 5.2

0.120

835

30.9 1

of the diluting fluid is much smaller than the extracellular space
of the rabbit and the additional outlet can, therefore, not play
any decisive role. The very rapid disappearance of potassium from
the circulation suggests the assumption that potassium, vhen
passing the endothelium, encounters appreciably less resistance
than does sodium or chlorine. The diffusion constant of potassium
in vater is larger than that of sodium; taking the former to be 1,
the diffusion constant of sodium makes out 0.65. The diffusion
constants of potassium and chlorine are practically identical. The
rates of penetration of potassium and chlorine through the endo-
thelium, however, differ greatly. The diffusion rate for water in
water was found, using heavw water as an indicator, to be 1.6
times larger, only, than that of chlorine or potassium, while the
rate of passage of water through the endothelium is very much

356

L. HAHN AND G. HEVESY.

faster than that of any other substance investigated by us. In
the course of 21 sec. the labelled water introduced into the circu-
lation of the rabbit is found to be distributed in 508 cc. body
water, corresponding to 34 percent of the rabbit’s weight.

“When investigating the staining capacity of dyes it was found
that generally the rate of coloration increases with decreasing
diffusion rate in water (Rous et alia 1930; Smith and Rous 1931;
Menkin and Menkin 1930). From the capillaries of the frogs
mesentery trypan blue was found to disappear exponentially
with a half-time period of 2 min. (Menkin and IMenkin 1939).

We have not yet mentioned the fact that an outlet of the plasma
®^Na, for example, is given by intrusion into the corpuscles. This
outlet is a very restricted one. We found the ®^Na content of 1 gm.
corpuscles of the rabbit to be, after the lapse of two hours, 11 per-
cent of that of 1 gm. plasma. From this figure and the haematocrit
value (34 percent) of rabbits blood it follows that, expressed in
diluting body fluid volume, the uptake of ®*Na by the corpuscles
corresponds to somewhat less than 4 cc., while the total diluting
volume of a rabbit weighing 2.5 kg. makes out as much as about
700 cc. After the lapse of V-jz hours, the content of 1 gm.
corpuscles was found to be 40 percent of that of 1 gm. plasma
and the content 48 percent. These figures correspond to an
additional diluting volume of 14 and 16 cc., respectively. The water
content of the corpuscles of the rabbit being about 63 percent, the
role of the corpuscles as additional outlet of the labelled water
molecules introduced into the plasma is not quite insignificant.

Permeability of Muscle and Brain Capillaries.

The figures stated in the preceding section give information on
the rate at which ions and molecules leave the capillary system.
Should a minor part of the capillary wall be slightly permeable
or even impermeable to some of the substances investigated, this
would not have been revealed by the figures given above, since
these figures indicate the permeability of the very inhomogeneous
capillary system in toto. If we want to know the average per-
meability of the muscle capillaries, for example, to ®*Na, we have
to compare the ®*Na content of. plasma and muscle samples of
known weight.

The results of such measurements are seen in Table 6, in which
the percentage ratio of the ®*Na content of 1 gm. fresh gastrocne-

RATE OF PENETRATION OF IONS.

357

Table 6.

Eatio of the ~^Na and content, respectively, of 1 gm. tissue and

1 gm. plasma.

Tissue

Time
in min.

Eatio of the content of 1
gm. tissue and 1 gm. plasma
X 100

32 p

Mnsclo

0.9

3.74

0.98

Muscle

5.2

7.75

3.42

Muscle

10.7

8.82

Muscle (other rabbit)

120

11.5

—

Brain, total

3.0

4.5

Brain, white

i 11

2.2

—

Brain, grey

! 11

3.9

—

Brain, white (other rabbit) ....

! 120

10.9

Brain, grey (other rabbit) ....

120

14.9

Medualla oblongata (other rabbit) . .

120

17.2

mius tissue to 1 gin. plasma is given. The table contains also
data on the «P content of muscles and the “Na and content of
the brain tissue. After the lapse of 11 min., a proportional parti-
tion of **Na between plasma and gastrocnemius is nearly reached,
since the size of the extracellular space of the gastrocnemius of the
rabbit is about 11 per cent of the weight of the muscles. The
muscle capillaries are seen to be more permeable to sodium than
to phosphate; in the course of the first minute about four times
more ®^Na left the plasma for the muscle tissue than ®“P.

The permeability of the brain capillaries to «Na and also to
®*P is lower than that of the muscle capillaries, those of the w ite
brain substance being apparently less permeable than those o
the grey brain substance to -*Na.^ After the lapse of 62 ours,
we found (Hahn et alia 1939) the =*Ha content of 1 gm. brain
to make out 32 percent of that of 1 gm. plasma, a figure w ic
corresponds to that stated (Manery and Hastings 1939) or

.sodium space of the brain. • -u i i n

Manery and Bale (1939) carried out experiments with labelie

sodium and phosphorus. They state that, in the course o an our,

‘ This is probably an expression of the fact that the vascularity of C gr y
matter greatly exceeds that of the "white.

L. HAHN AND G. HEVESY.

358

the uptake by the brain and the sciatic new is much smaller
than to be expected in the case of a proportional partition of
between plasma and interspaces of these organs while, in the case
of the other organs, such a partition was reached after the lapse
of 20 min. Wallace and Brodib (1937, 1939) investigated the
distribution of iodide, thiocyanate and chloride in various tissues
of the body and found that the relative concentration in terms of
blood tissue ratio was alike for the three substances in the organs
examined, with the exception of the brain in which the chloride
was in much larger amounts that the other two. In a later in-
vestigation (1939) these authors found that these anions distribute
in the central nervous system in the same ratio to chloride as in
spinal fluid, whereas in other body tissues they distribute in the
same ratio to chloride as in serum. That sulphate passes more
slowly than bromide or nitrate through the brain capillaries can
be concluded from experiments in which the chloride content of the
plasma was replaced by sulphate (Ambeeson et alia 1938), bromide
(Weie 1936), and nitrate (Hialt 1939), respectively.

While the capillary wall in toto was found to be more per-
meable to phosphate ions than to sodium ions the opposite is the
case for the wall of muscle capillaries. This fact suggests that a
substantial part of the capillary system, possibly that of the bone,
must be very easUy permeable to phosphate.

We compared, furthermore, the activity of 1 gm. of the grey
brain substance with that of 1 gm. plasma 59 min. after the
administration of radiobromide. The ratio found was 9.3 per cent.
As the chlorine space of the brain tissue of the rabbit is 35 per
cent, we have to follow that in the course of an hour, less than of
the proportional partition of bromine between plasma and the
extracellular volume of the brain is obtained.

Permeability to Water.

We investigated previously the rate at which heavy water
introduced into the circulation leaves the plasma (Hevesy and
Jacobsen 1940). We extended these measurements by deter-
mining the distribution of labelled water between plasma and
gastrocnemius of equal weight. The figures obtained are seen
in Table 7. The water content of the samples was driven off and
was collected in vacuo as described previously. We are much
indebted to Miss Hilde Levi for kindly determining the density

KATK OF PENETRATION OP IONS.

359

Fig. III. Percentage distribution ratio of labelled sodium, potassium and deuterium
oxyde between plasma Trater and muscle water of equal weight.

Table 7.

Ratio of the dencity excess of plasma water and muscle water.

Time in min.

Ratio of density excess of plasma water and
muscle water

1.62

1.16

1.06

1.00

of the water samples obtained by making use of Linderstebm

Lang’s floating drop method. i f

The density excess of these water samples over the density ot
normal body water is a measure of their labelled water con en .
The ratio of the density excess of plasma water and muse e wa er is
stated in Table 7 and Figure III.

360

L. HAHN AND G. HEVESY.

A ratio equal to 1 is to be expected wben tbe heavy water con-
centration of the total muscle water corresponds to that of the
plasma water. This stage of partition is reached between 20 and
38 min. after the start of the experiment.^ This is the same result
at which Hevesy and Jacobsen (1940) arrived when investigating
the rate of disappearance of heavy water from the circulation.
The water content of the gastrocnemius makes out 77 per cent of
the muscles weight, about 11 per cent being located to the inter-
spaces and the rest to the cells. If the extracellular water would
alone take part in the exchange process, we should expect in the
case of a proportional partition of the heavy water between plasma
and extracellular water the ratio of the density excess to be about
7. From the fact that this ratio is found after the lapse of 5 min.
to be but 1.62 we have to conclude that during that time not only
a proportional partition of the labelled water between plasma and
the extracellular fluid of the muscles took place but a large part
of the cell water was replaced by plasma water as well.

Summary.

Solutions of labelled chloride, bromide, sodium, potassium,
phosphate and deuterium oxyde were injected into the circulation
of rabbits and the speed of the escape of the labelled ions from
the plasma was determined. Potassium was found to penetrate
at a much faster rate through the capillary wall than any other
ion investigated.

Information on the permeability of the muscle and brain ca-
pillaries were obtained by comparing the labelled sodium, phosph-
ate and heavy water content of muscle and plasma, respectively
brain and plasma. The muscle endothelium was found to be more
permeable to sodium and to phospate than the endothelium of
the brain. The partition ratio of labelled sodium between plasma
and the chloride space of the brain amounts after the lapse of
11 min. only to Yio of the equilibrium value; during that time a
proportional partition of labelled sodium between plasma and the
chloride space of the muscle was obtained.

In the course of an hour, somewhat less than of the propor-
tional partition of bromine between plasma and the grey brain

* Though great care was taken to diatill off the total water content of the
muscle, the possibility that a minor amount of muscle water was not removed can
not be excluded. This water may not have taken part in an exchange process.

RATE OP PENETRATION OF IONS.

361

substance was reacted. Proportional partition of labelled water
between plasma and tbe muscle water was reacted after about
talf an tour.

We wist to express our tearty ttanks to Professor Niels
Bohr for kindly putting numerous facilities at our disposal.

Eeferences.

Aaiberson W. B.., T. P. Nash, A. G. Mulder and D. Binns, Amer.
J. Ptysiol. 1938. 122 , 221.

Griffiths, I. H. E., and B. G. Margbaith, Nature, i939, 143 , 179.
Hahn, L. A., G. Gh. Hevesy and 0. H. Kebbe, Biochem. J. 1939, 33 ,
1549.

H!E\msy, G. and C. F. E. Jacobsen, This Archiv 1940. 1 . 11.

Hialt, E. P., Amer. J. Ptysiol. 1939. 126 , 553.

Keys, A., Proc. Faraday Soc. 1937. p. 932.

Krogh, a.. Anatomy and Physiology of the Capillaries, New Haven
1929.

— , Acta med. scand., 1937. Supplement XC.

La\tetes, P. H., j. Bourdillon and K. A. Klinghoffer, J. din.
Invest. 1936. 15 , 261.

Manery, j. F. and W. F. Bale, Amer. J. Ptysiol. 1939. 126 , 578.
Manery, j. F. and A. B. Hastings, J. biol. Ctem. 1939, 127 , 657.
Menkin, M. and M. F. Menkin, J. exp. Med. 1930. 51 , 285.

Rous, P., A. P. Gilding and F. Smith, Ibidem. 1930. 51 , 807.

Smith, F. and P. Rous, Ibidem. 1931. 53 , 195.

Wallace, G. B. and B. B. Brodie, J. Pharmacol. 1937. 61 , 397 and 412.
— — , Ibidem. 1939. 65 , 214 and 220.

Weir, E. G., Doctoral thesis. University of Chicago 1936.

From the Women’s Clinic, University, Lund (Prof. A. Westman),
and the Women’s Clinic, General Hospital, Malmo (Dr. S. Gekell),

Sweden.

Studies on the Muscular Physiology
of the Genital Tj’act.

n. Tonus, Spontaneous Activity and Drug Beactions
in the Cervical Muscles/

SUNE G-ENELL.

(With 6 figures in the text.)

In the extensive research- work devoted to the physiology of
the uterine muscles (review in Genell, 1940) almost no atten-
tion has been given to the' cervical muscles. Yet the anatomical
situation of the cervix as an intermediate link between vagina
and uterus is in itself a reason for assuming that its muscular
apparatixs has an important function. The present investigation
has been carried out on rats.

Tonus.

In a series of experiments for another purpose (Genell, 1939)
the observation was made that an injection-needle furnished with
a 1-mm broad bulb passes through the cervical canal of a non-
narcotized animal with ease when the animal is in dioestrus, but
with difficulty when it is in oestrus. This observation suggested
the thought that the tonus conditions in the cervical musculature
differed from those in the rest of the uterine muscles.

Method.

The calibre of the cervical canal was measured by means of a
series of measuring-cylinders resembling miniature Hegar dilators
and graduated in sizes of 0.5, 0.8, 1, 1.5, 2, 2.5, and 3 mm. An
assistant held the non-narcotized animal. As vaginal speculum an

* Received 15 November 1940.

THE MUSCDLAR PHYSIOLOGY OF THE GENITAL TRACT. n{)n

ordinary funnel-shaped ear speculum of suitable size u'as used. A
head-lamp or head-mirror facilitates the ^vork. After the portio
vaginalis had been brought into the speculum, the vagina was
dried up and the measuring-cylinders passed through one of the
cervical canals. If a certain cylinder did not pass smoothly
through the canal, the next higher size was cautiously tried.
If this could not be passed through without undue pressure, the
thickness of the preceding cylinder that could was taken as the
value of the canal calibre.

Experiments 1 — S. Eight animals with a normal sexual cycle were
followed from the pro-oestrous phase (first day, 1 p. m.) to the met-
oestrous iihasc (second day, 7 p. m.). Guided by changes in the vag-
inal smear the calibre of the cervical canal ^Yas measured at different
periods (Table 1).

Table 1.

1 Stage*

i ■ ”

; Phase'

P 1

Pro j

P 0 s

7 ; S j 9 11 , 12 1 1

Met

jTimc of
j day

1 f)

p. m.

ID.

p. m.

■Exp.| Rat

Calibre of cervical c.anal in mm at above periods

: 1 ;824i O.r.

1 1

j 0.8 1.0 j

1.5

: 2 :S41

0.8 ;

. O.s l.O i

i ;

1.6

3 j849

0.8 j 1.0 ■

1-0 1 '

1.5

0.8

1-0 . ; ■

1.5

.5 I860

l.O 1

j 0.8 j 1.0 j

1.5

6 jS68

0.8 j ;

1.0, |l.o,

1.5

7 1 873

0.8 i

i 1.0 i ;

1.0

il.5

1.5

; 8 }874

0.5

‘ i 1.0 1 • l.O'-

1 \ \

1.5

' 1.5

Table 1 shows that in the pro-oestrous phase the cervical ca-
nal is only in exceptional cases (Exp. 5) greater in calibre than
0.8 mm, at times less. The canal gradually increases in calibre,

^ >Stage> and >Phase> — for explanation see Gendl: Acta physiol, scand.
1940: 1; 139.

25 — 'i01323. Acta phi/s. Scandinav. Vol.I.

364

SDNE QENELL.

SO that during the first half of the oestrous phase (pos = Stage
II) it measures about 1 mm. In the later half of the oestrous
phase (= Stage III) it increases further to 1.5 mm. In the sub-
sequent course of the sexual cycle (only partly recorded in the
table) the cervical canal allows the passage of at least a 1.5 mm
cylinder, sometimes a 2 mm one.

The cervical canal is invariably narrow during the whole of
the pro-phase and beginning of the pos-phase, i. e. during a period
of about 24 hours. It seems justifiable to interpret the varia-
tions in the calibre of the cervical canal as tonus variations in the
unstriated muscle.

Experiments 9 — 14. Ten castrated animals. The calibre of the cer-
vical canal was measured. Varying doses of oestrin were adminis-
tered to the animals and the calibre of the cervical canal re-measured
(Table 2).

Table 2.

Exp.

Eat

Cervical Canal:

Oestrin treatm.

Cervical Canal; calibre

Day

after

castr.

Calibre

in mm

Day

after

castr.

Dose
in I. U.

Day

after

castr.

Vaginal smear

in mm

944

neg;*pro

1.5

946

neg-pro

1.5

952

l.s

(pro)

1.5

961

(pro)

1.5

963

neg

969

pro

944

150

(pos)

1.0

949

1.5

150

p^o-po^

1.0

952

1.6

150

pro

0.8

961

1.5

0.6

963

1.6

300

0.8

969

1.6

300

. 16

pro

0.8

Table 2 shows that 7 — 14 days after the castration the calibre
of the cervical canal is 1.5 mm. Oestrin doses of 16, 30 and 45

TUE MUSCULAR PHYSIOLOGY OF THE GENITAL TRACT. 365

i. H. have no constrictive effect on the cervix, while doses of 150
i. u. reduce the calibre a little. I^ot until a dose of 300 i. u. (which
has been experimentally shown to be the adequate oestrogenic
dose for the rat, Gexell, 1937) has been reached, does the typi-
cal narrowing of oestrus to a calibre of 0.5 — 0.8 mm appear in the
cor\'ical canal.

From these experiments it can be seen that the tonus of the cer-
vical muscles of the cervix is high during oestrus and low during
dioestrus and in castrates. Treatment of castrates with an ade-
quate oestrogenic dose of oestrin brings the cer\dcal muscles in-
to the tonic state characteristic of heat. The variations in the
tonus of the cervical muscles during the sexual cycle are thus
regulated by the oc.strous hormone, oestrin.

Spontaneous Activity and Brug Reaction.

Fourteen animals have been tc.sted. As the cervical musculature
of the rat is almost exclusively circular, the cervix prepara-
tion was made in the following manner. The uterus and vagina
were cut completely away. All the connective tissue of the para-
metrium was removed. The cervix prc-dissected in this way con-
sists of a firm, conical body 7 to 8 mm long and 3 — 5 mm wide.
It was sliced transversely into about 1-mm thick rings. At the
two lowest of these the cervical canal is still single, at the upper
it has forked. For recording purposes the lowest but one (N) and
the uppermost but one (0) of these rings were selected. The cus-
tomary dispositions for the Magnus-Kehrer experiment were
adopted. Medium: 100 cc scrum saline solution: NaCl 8, KCl 0.42,
CaClj 0.24, MgCl. 0.005, NallCOj 1, glucose 0.5 gra per 1000 cc.
Oxygen with 5 per cent CO. was bubbled through. Of the 14 ex-
periments six are selected for reproduction /Figs. 1 — 6/.

Spontaneous Activity. In surviving preparations the circular
muscle of the cervix, like the muscle of the uterine cornu,
develops a spontaneous rhythmic activity. This shows a slightly
greater contraction frequency and a slightly greater amplitude in
the upper part of the cervix than in the lower. The differences
betw'een the sexual phases are less striking in the circular muscle
of the cervix than in that of the cornu uteri.

In pro-ocstrus (pro. Figs. 1, 2, 3) the rh 3 d;hm seems to be some-
what more rapid than in oestrus (pos. Fig. 4, lower curve), met-
oestrus (met, Fig. 5) and dioestrus (neg 4, Fig. 4, upper curve). In

366

SUNE GENELL.

IW ■ 1 1. '■■ i.MH^ l.l.l 1.1 1.»<.1.».».».».».^.».«.-^.».'W.^.W.^.^. .A^ (V ^ IV f V

6^2., ~^y. "p** , er B 6

JuiiiiiliuuiiailiiiiummMMmmmmmmmmmmMmmmmMmmmmmimmmiimmmmm

Fig. 1. Rat, pro-oestrns, cervix, circular musculature, upper (6) and lower (N)
segment. Pen-stroke ratio 1 : 7. Medium 100 cc. Time in min.

R« ^? 6 , 'Jd?, -foociSo

tuiii„imimi».iiiiiiimii.i,iiimiiuiiiilll.....i..uillllll, I, im...

Fig. 2. Rat, pro-oestrus, cervix, circular musculature, upper (0) and lower (N)
segment. Pen-stroke ratio 1 : 7. Medium 100 cc. Addition 100 y adrenaline.

Time in min.

^0 m^'T ^CL

L’fitfitfWiWWiWWWi

Fig. 3. Rat, pro-oestrus, cervix, circular musculature, upper (0) and lower (N)
segment. Pen-stroke ratio 1:7. Medium 100 cc. Addition (from left) O.I, 0.6
unit pitocin, 10 Mgm BaCl,. Time in min.

the three last-mentioned phases, moreover, the spontaneous ac-
tivity thins out (and ceases) far quicker than in pro-oestrus. The
difference between oestrus and dioestrus in the cervical muscles
is accordingly the reverse of that observed in the rest of the
uterine muscles of the rat (Genell, 1936, 1937).

THE MUSCULAR PHYSIOLOGY OF THE GENITAL TRACT. 367

Castralion brings about a profound change in the spontaneous
actmty. In the upper cervix: this activity almost ceases, in the
lower it ceases altogether (Fig. 6).

Drug Reactions. The cervical muscles show no response to -pi-
locin (Figs. 3, 4), either by increase of contraction frequency and
amplitude or by increase of tonus. In pro-oestrus (Fig. 3) an im-
doubted increase in the frequency of contraction is observed after
a (unphysiologically) high dose of pitocin (^i unit per 100 cc).
Adrenaline has the same effect on the cervical muscles as on the.
rest of the uterine muscles; inhibition and depression of tonus
(Fig. 2). Bad. has the same effect on cervical muscle as it has
on all other unstriated muscle (Figs. 3, 4, 5, 6).

Discussion.

"Whereas the tonus of the rat’s uterine muscle (Genell, 1940)
is high in the dioestrous phase and low in the oestrous, the
reverse applies to the cervical muscles, as can be seen from
the present investigation. Substitute experiments on castrates
show that the change of tonus is evoked in both cases by
oestrin. Thus, as regards muscle-tonus, the oestrous hormone
of the ovary produces contrary effects in the cervix and corpus
uteri. The spontaneous actmty of the rat’s uterus (corpus) is
characterized by low frequency and great amplitude in the oest-
rous phase, by high frequency and small amplitude in the dioest-
rous phase (Genell, 1936, 1937). The reverse state prevails in
the cervical muscles. Ne\vton (1933), worldng on surviving
preparations from guinea-pigs and rats, showed that cer^^cal
muscle is refractory to the posterior-lobe hormone. This fact is
also brought out by the present investigation. A condition con-
trary to that in the muscular apparatus of the corpus uteri is also
present here. On the other hand, the cervix does not respond to
adrenaline in the same manner as the rest of the uterine muscles.

Siunmary.

In afro-investigations of the tonus conditions of the cervical
muscles have shown that the tonus is highest in the oestrous
phase, lowest in the dioestrous phase and after castration. In
substitute experiments it has been shown that the increase
of tonus in oestrus is elicited by oestrin. Surviving prepara-

368

SDNE 6EKELL.

. — ,vOV/LaA/l , ..yvA>\/v ^ . . AA.a

t( ^ ^ ^ 4 f

ioo

/VO

/*o

m^lJMlUUUUUUUlAJL

^4 4- 4

J\ (L

Wiiiiii.iii.iiii,m,ii,iiiui„ii.„,ii„iiiiiii,iiii,iiii,iiii„„i„„„i„,.

niliiiiiim

Fig.*4. Upper curve: rat, dioestms, cervix, circular musculature. Lower
curve', rat, oestrus, cervix, circular musculature. Pen-stroke ratio 1:7. Me-
dium 100 cc. Additions (from left) O.Ol, 0.02, 0.02, 0.06, 0.1, 0.8, 0.5 unit pitocin,
10 Mgm BaClj. Time in min.

MISMIMWMJJULJL^ _jiiiAiiiUUliliUU

BBifiSJL/UUUU'LJl

A ti ^ ^ n ^ t

'* 0

fo o Cc 80

Fig. 5. Rat, metoestrus, cervix, circular musculature, upper (0) and lower (N)
segment. Pen-stroke ratio 1 : 7. Medium 100 cc. Addition 10 Mgm BaCJ,.

Time in min.

I.'

f?€L G^7, ^^7, fCa,sfr. 9

f
1 7.

'foo cc 80

Fig. 6. Eat, castrated 9 days earlier, cervix, circular musculature, upper (0)
and lower (N) segment. Pen-stroke ratio 1 : 7. Medium 100 cc. Addition
10 Mgm BaClj. Time in min.

tions of circular muscle from tlie cervix have been studied in
different sexual phases. In the oestrous phase the spontane-
ous activity is of quicker rhythm and greater duration than in
the dioestrous. After castration the spontaneous activity tends
to cease.

Pitocin is without effect on the cervical muscles.

THE MDSCULAR PHYSIOLOGr OF THE GENITAL TRACT. 369

AdrenaUnc causes inliibitiou and depression of tonus in the
cervical muscles.

The cervical musculature, in respect of spontaneous actmty
and tonus, behaves inversely to the rest of the uterine muscida-
ture, and in contrast to this is refractory to pitocin but like this
is negatively adrenotropic.

References.

Genell, S.; Acta obstet. g3mec. scand. 1936. 16. 51.

— , TJterin- och vaginalmuskiilaturens funktionella forhallande i den
icke-g^a^^da organismcn. Lund, 1937.

— , Acta obstet. gynec. scand. 1939. 19. 113.

— , Acta physiol, scand. 1940. 1. 139.

— , Ergebn. physiol. 1940. 43. 371.

Newton, W. H.: J. Physiol. 1933. 79. 301

From the Neurophysiological Laboratory, The Caroline Institute,

Stockholm.

Rotation of Activity and Spoil tiineous Rliytlinis

in the Retina.^

RAGNAR GRANIT.

(tVitli 4 lignres in the text.)

Several years ago (GRA^'IT and Therman, 1935) we made the
first observations on rotation of activity in the eye. Later this
same phenomenon turned up in some experiments with constant
stimuli as an alternation between types of electroretinograms so
different, that we felt compelled to speak of a “stvitchboard” in
the retina (Granit and Mtjnsterhjelm, 1937). With rhythmic
stimuli rotation of activity has also been observed by Bartley
(1937) and recently Bartley (1939) and Berger and Buchthal
(1938) have emphasized the general significance of this phenom-
enon in the physiology of vision.

When I now return to this subject it is partly because our pres-
ent micro-electrode technique (Grakit and Svaetichin, 1939)
brings it out with hitherto non-paralleled clarity and regularity,
partly because of my conviction that rotation of activity is a very
essential property of the nervous system.

From the point of xdew of the general physiology of the special
senses one would be prepared to state that there are good reasons
for accepting as a law of end-organ differentiation that the evolu-
tion of structurally more complicated sense organs takes place
in such a manner as to counteract the effects of adaptation with
their tendency to make the stimulus ineffective. Rotation of
activity is one of the means whereby tliis end is attained. The
retina which is forced to continuous, accurate activity within an

^ Received 2 January, 1941.

ROTATION OF ACTIVITY.

371

enormous range of illumination has this mechanism well developed
and, in addition, possesses several other means of suppressing or
modihung the effects of adaptation. These, however, cannot be
discussed now.

Technique and Procedure.

The experiments were carried out with the silver-micro-elec-
trode described by Granit and Svaetichin (1939). In many cases
the glass-covered platinum micro-electrodes of Taylor and
Whitaker (1928) were used, yet without making them non-
polarizable. Experiments proved, as was to be expected, that the
metal-micro-electrode has a polarization capacity which greatly
deforms the slow' retinal action potential without preventing the
fast spikes from appearing. If the glass covered metal-micro-
electrode be used with a directlj’^ coupled amplifier it is found to
record a rectangular input current with the distortion charac-
terizing a condenser-c.oupled instrument having somewhat larger
coupling condensers than those used in my amplifier for micro-
recording of spikes from the retina.

The micro-electrode is applied on to the retina with the aid of
a micro-manipulator under a binocular microscope. Leads are
taken to a condenser-coupled amplifier with a balanced input
stage. A cathode ray and a loudspeaker are connected to the out-
put stages in the usual manner. When the light is switched on or
off discharges composed of spikes follow' from elements w’hich are
more or less isolated as the case may be (see figures). These remind
one of the spikes recorded with micro-electrodes, for instance, by
Forbes and his collaborators Renshaw, Thermae etal. (1937,
1940, 1940) from the hippocampus area and by Lorente De No
(1939) from the nuclei of ocular nerves.

Gr.anit and Svaetichin (1939) held these spikes to arise from
the neurite not too far from the axone hillock of the retinal gang-
lions. The reason for this, not mentioned in their paper, was
that spikes could be obtained when their micro-illuminator was
pushed into the retina several millimeters away from the micro-
electrode. According to Hartline (1939) the "receptive field”
of a single fibre would have much narrow’er dimensions and the
retina itself is only a fraction of a millimeter. On the other hand
the spikes cannot arise too far away from the ganglions as they
are absent or minute in the blind spot.

372

RAQNAR GRANIT.

In the experiments to be described below the ivhole retina was
illuminated with some wave-length from our monochromator.
For technical details the reader is referred to the paper by Granit
and SvABTiCHiN (1939). The experimental animals were frogs
and tortoises.

Results.

The Response to Intermittent Stimuli.

Rotation of activity is conspicuous and most disturbing in
experiments presupposing a constant tlireshold. To this type be-
longs the colour work taken up by Granit and Svaetichin [1939)
and at present continued in this laboratory with a variety of
animals. Sometimes one finds the threshold to undergo sudden
changes which coidd be ascribed to the pressure of the micro-
electrode were it not for the fact that these changes may come
and go without, in the long run, involving any progressive di-
minution of the sensitivity. In fact, I have followed dark-adapta-
tion for two hours with a well placed micro-electrode. That pres-
sure can lead to spontaneous discharges is a different matter. But
this source of error can with some experience be avoided and then
the sudden shifts in the level of sensitivity must be ascribed to
causes in line with those leading to the spontaneous rhythms to
be described in the next section. These factors cannot, as a rule,
be put under experimental control, though the phenomena are
interesting to record and try to modify when they occur.

But inasmuch as rotation of activity depends on stimulation,
which may be supposed to activate after-potentials or other
mechanisms of blocking or facilitation, then one certain way of
regularly bringing these into play would be intermittent illumina-
tion. Every flash of light then leads to a state of excitation
which has to force its way through a bed of receptors and neurons
modified in excitability by the foregoing flashes. Records, from
such experiments are shown in Fig. 1.

From above downwards I have selected cases designed to illu-
strate the rotation of activity of a gradually increasing number
of neurons. There are also variations in the frequency of the
flashes and in their strength. The essential features of the phe-
nomenon are displayed, independently of the conditions chosen.
The units come and go in a rotation of activit}’ w'hich is irregular
w'ith respect to the rhvthm of stimulation. The active unit of

Fig. 1. The response to intermittent light at different frequencies. Light signal
above time signal (50 per sec.) in this and the following records.

1. Tortoise, wave-length O.CSO “Off”-spikes.

2. Frog, wave-length not noted. •

3. Frog, spikes at both “on” and “off”. I^ear threshold for wave-length 0.530 «.

4. Tortoise, wave-length O.020

5. Frog, wave-length O.COO Well above threshold for intermittent light.

cuive 1 pauses every nov and then, and again returns after a fevr
flashes. In curve 2 there are three units fairly well placed relative
to the electrode and the picture in this case as well as in curves
3 — 5 vdth a greater number of elements in actitiUty is already
very complicated. Continued stimulation with intermittent light
neither seems to make the rotation of activity more regular nor
does it abolish it.

Spontaneous Rhythms.

ADElAi? and Matthews (1928) made the first observations on
spontaneous rhythms and sjTichronization in the retina when re-
cording from the whole nerve. The off-effect is quite often synchro-
nized (Granit and Thermak, 1935), particularly in the fibres that
merely react to cessation of illumination (Hartlike, 1938).

With the micro-electrode technique spontaneous discharges
are quite common and very often those consist of grouped spikes

374

BAGNAK GRANIT.

'*'' ♦ ' '4H

x^. in . I.» . nl tf f»jt|Ktii|^iii^iiiii>^ M t I i m ■ I )| || I II I I ■■.

Fig. 2. Tortoise. Upper curve: off-effect control.

Lower curve: inhibition caused by re-illumination. Wave-
length O.GOO ft, near threshold.

■which for some time may be synchronized. In connection with
the problem of rotation of activity it interested me to find out
to what an extent such discharges and rlij^thms, when they occur,
can be modified by stimulation.

Very variable, though in each case repeatable, results were
obtained. At one end of the series of observations could be jdaced
I the ordinary inhibitable off-effect, illustrated in Fig. 2, Some-
what similar properties has the spontaneous discharge of the darlc-
adapted eye, particularly when it is diffuse and not synchronized.
With the frog’s eye it is often a sign of dark-adaptation that the
retina begins to discharge spontaneously. This discharge, as a
rule, is very effectively inhibited by illumination. "Dark” rhythms
of the eye of the water-beetle {Dytiscus marginalis) behave simi-
larly, according to Apriax (1937).

An interesting case is curve 1 of Fig. 3. There was a spontaneous
discharge ha'ving the fairly regular rhythm illustrated. A light
flashed into this discharge (see the curve) led to a temporary in-
hibition probably caused by the on-discharge elicited in the
neighbourhood. Signs of this on-effect are seen under the micro-
electrode. Cessation of stimulation was followed by a brief off-
discharge and another temporary silent period after which the
receptor again picked up its original rhythm. The rhythmic dis-
charge lasted for some time so that the experiment was several
times repeated. The results show that cessation of illumination
also can exert inhibitory activity. This hitherto seemed to be a
prerogative of illumination falling into an off-discharge. The

ROTATION OP ACTIVlTy.

. 1 ^ 1 . i . . .. L..X..L .l..l I . 1 . I 1..1 iliL

■I ' I. I ' r 1 I . . f , r-f.-

... j . ,

III I

»*«{iw< 4 i

fSA^W^AA^^^XiK^^^AA/WAAAi^^iA.^AAAAAAAA.^iVA.AA^AAAAAAA./WVAAAAAAAA/V^^^NV'

;15.. I...

' ' [ i

> — t > . ...1

-p —

1 ■ 1

1 ^ ^

i . } 1

I T .

Fig. 3. 1. Tortoise. Spontaneous rliythm inhibited by onset and cessation of
stimulation with wave-length 0.020 u at maximal intensity of monochromator.
2 and 3. Tortoise. IVave-length 0.G20 u, ^^•ell above threshold (see text).

4. Tortoise. Active unit firing spontancouslj' also during illumination.

Note. The single units illustrated in these experiments are the “red” receptors
of the tortoise with maximal sensitivity around O.G20 /‘ (Gbanit 1941).

inhibition at “off” is probably to be parallelized -with the well-
knoTvn “silent periods” of other receptors.

The counterpart to curve 1 of Tig. 3 is curve 4. This shows the
spontaneous activity of a single element which seemed to be
quite independent of Avhether the eye was illuminated or not.
This is nearly always the case vuth a discharge caused by the pres-
sure of the micro-electrode. But here there was no reason to
suspect that mechanical stimulation had caused the activity.
AVhen slowly screwing the micro-electrode into position and
looking at it all the time in the mieroscope one often first hears
the rhythm in the loudspeaker as a faint distant noise which then
gradually increases in strength as the point of the electrode
approaches the active unit from above. "lAHien this is so there is
no reason to ascribe the discharge to mechanical stimulation, the
less so as the spontaneous activity quite often may cease as
suddenly as it has begun. 'N^Tien light does not influence such
rhythms the reason may have been that the monochromator gave
too feeble stimuli. But there was always a check on this in the

376

RAGNAE QRANIT.

jr- "W^ 'V' w^ W' ' ' ''‘ 1 ^ 1 ^*

M WW» M M lli>

Fig. 4. Frog’s e^'c which has been in the apparatus for 1 h. 40 min. During this
time it has dark-adapted and begun to discharge spontaneously 15 — 18 times a
minute. The active region is of the type giving a “grouped discharge”, which comes
above and goes below the threshold as a grouped unit when the intensity is varied.
The spontaneous discharge looks like the discharge caused by stimulation. Wave-
length 0.520 If. For explanation, see text.

ab. 2,400 m. c. light of the lamp of the preparation microscope
illuminating the retina from above. This light could be switched
on and off as a final control on the relative independence of the
rhythm.

There are undoubtedly, particularly in the eye of the tortoise,
spontaneous rhythms which behave as if they had succeeded in
blocking every path around the discharging units.

Curves 2 — 3 of Fig. 3 show the beginning and end of inter-
mittent stimulation with the micro-electrode in a place which at
the moment of recording was silent, but some minutes later began
to discharge spontaneously. At the moment of recording the ten-
dency to spontaneous activity probably was in statu nascendi and
its appearance was facilitated by the flashes from which the
rhythm of the discharge rapidly disconnected itself.

In such cases the result may become very complex. Kg. 4
illustrates an eye tending to give a grouped, spontaneous dis-
charge Very much like the ensueing response to flicker. The ac-
tive spot is silent when the recording begins and, upon illumina-
tion, follows the slow rhythm of the intermittent stimulus. Spon-
taneous activity begins soon (end of curve 1) but immediately is
pressed into the rhythm of the stimulus where (curve 2) the extra

ROTATION OP ACTIVITY.

oil

discharge takes up a definite place. In curve 3 it has disappeared
and the eye now again reacts as in the beginning (curve 1).

These experiments, chosen to illustrate some properties of
spontaneous rhythms in the retina, could easily be multiplied
and would then provide samples of most of the phenomena of a
similar nature observed in the central nervous system. But as
they seem to have relatively little analytical value in their pre-
sent form I have merely used them as a complement to the observa-
tions on the rotation , of actmty. From this point of view their
significance is that they illustrate how inhibition can block a dis-
charge of spikes and how rhythms imposed by a stimulus can play
upon rhythmic tendencies in a given group of neurones.

Discussion.

It is clear that excitation, spontaneous or caused by stimula-
tion, is surrounded by inhibitory and excitatory influences
spreading over the retina in complicated patterns. These may
well be of the nature of after-potentials. The net result of the
waxing and waning of such effects must lead to rotation of ac-
tivity among the active elements. There is also at the threshold
a fluctuation of excitability which may or may not be of different
origin, but whether it is of any significance mth “flicker” is open
to doubt. In order to make a group of spikes follow an inter-
mittent stimulus the illumination must be well above the thres-
hold for a constant light.

For the physiology of ^'ision the rotation of activity, apart
from what it may do to counteract the affects of adaptation, is
of great interest and emphasizes points' of view advocated by
Bartley (1939), Berger and Buchthal (1938) and Wright and
Geanit (1938).

Summary,

“Spikes”, recorded from the retina with micro-electrodes and
a condenser-coupled amplifier, are elicited by stimulation with
intermittent light.

Single units of activity as well as a greater number of active
elements responding to intermittent light, show a very marked

378

BAGNAR GBANIT.

rotation of activity, the individual elements pausing and re-
entering into activity at irregular intervals.

Spontaneous activity in the retina can with respect to an
interposed stimulus be dhuded into two categories, (i) discharges
which can be temporarily or completely inhibited by illumination
or even temporarily inhibited by cessation of illumination, and
(ii) discharges which continue independently of whether the eye
is illuminated or not.

Spontaneous rhythms can be activated and facilitated by rhyth-
mic stimulation.

The experiments thus give further support to the view that
waves of excitation passing through the retina are surrounded by
spreading patterns of inhibitory and excitatory influences. With
stimuli well above the threshold, these probably are the main
causes of the rotation of activity.

From the sensory point of view rotation of activity may be
regarded as one of the means whereby the effects of adaptation
are counteracted, and it also emphasizes the fundamental truth
behind "pattern” theories of visual processes.

The experiments have been carried out with the support of a
grant from The Rockefeller Foundation to this laboratory.

Beferenoes.

Adrian, E. D., J. Physiol. 1937. 91 . 66.

— and R. Matthews, J. Physiol. 1928. 65 . 273.

Bartley, H, S., Amer. J. Physiol. 1937. 120 . 184.

— , Psychol. Rev. 1939. 46 . 337.

Berger, C. and F. Buchthal, Skand. Arch. Physiol. 1938, 79 . 15.
Forbes, A., B, Renshaw and B. Rempel, Amer. J. Physiol. 1937,
119 . 309.

Granit, R,, Documenta Ophthal. 1938. 1 . 7.

— , This Journal. 1941. 1 . 370.

— and A. Munsterhjelm, J. Physiol. 1937. 88 . 436.

— and 6. Svaetichin, Upsala Lakaref. Forhandl. N. F. 1939. 45 .
161.

— and P. 0. Therman, J. Physiol. 1935. 83 . 359.

Hartline, H. K., Amer. J. Physiol. 1938. 121 . 400.

— , Amer. J. Physiol. 1939. 126 . 527.

Lorente Be No, R., J. Neurophysiol. 1939. 5 . 402.

Renshaw, a., a. Forbes and B. R. Morison, J, Neurophysiol. 1940.
3 . 74.

ROTATION OP ACTIVITV.

379

Taylor, C. V. and D. M. Whitaker, Protoplasma. 1928. 3 . 1.
Therman, P. 0., A. Forbes and R. Galambos, J. Neurophysiol. 1940.
3 . 191.

Wright, W. D. and R. Granit, Brit. J. Ophthalm. Monogr. Suppl. IX.
1938.

26 — 401323. .Icfa phys. Scandinav. Vol. I.

From The Neurophysiological Laboratory, The Caroline Institute,
•Stockholm, and The Anatomical Institute of
Helsingfors University.

Dark- Adaptation and the Platinum Cloride
Method of Staining Visual Purples

STEN STENIUS.

(With 1 Fig. in the text.)

From different points of view Granit et al. (1938, 1939) and
Lythgoe et al. (1936, 1940) have arrived at the conclusion that
visual purple acts along the surface of the rods where, accord-
ingly, its concentration should be maximal.

This conclusion formed the starting point of my attempt to
try Sterx’s (1905) platinum chloride method for the staining
of visual purple in order to find out how the substance is distri-
buted. Stern’s original description does not give the details
of his procedure, but in a personal communication to Prof. Gra-
NiT some valuable suggestions were made by Dr. Tansley who
also has used this method (1933). Some 40 — 50 eyes were spent
in my own work and the procedure that finally emerged from
these trials led to good and repeatable results. Description of
this method should save a great deal of labour for anyone Avant-
ing to apply it.

The frogs {JRana escidenta) were dark adapted and decapitated
in red light, the eyes excised and freed from all slices of adjoin-
ing tissue. For each eye 2 cc of a 2*/^ % platinum chloride so-
lution were used. The best method proved to be injection of
the fixative through the optic nerve into the bulb Avdth the aid
of a fine needle.

After this the eyes were left for not less than 12 hours in the
platinum chloride solution. In this manner bleaching of the
visual purple was prevented before fixation had taken place.

^ Received 17. November 1940.

Fig. 1. Microphotograplis of cross sections of frog retinae in the dark-adapted
state fixated (a) in plationm chloride, (b) in Zenker’s solution. Zeiss ultrnphot.

Then the eyes were cut open and the anterior half of the bulb
removed. Then followed fixation in absolute alcohol for 1 hour,
the fluid being renewed after half an hour. The next step was
clearing in xylol for half an hour, the fluid being renewed after
15 min. Finally the eyes were .soaked for one hour in melted
paraffin (52°) and embedded in the paraffin of a second bath
(58°) where they also had been for one hour.

Sections at 10 /r did not change their colour during the first
few days if they were kept in the dark. In daylight the yellow
colour gradually disappeared (but see Stern, 1905).

Of greatest interest proved to be the cross sections of the rods
illustrated in fig. la. Fig. lb is a control fixated in Zenker’s
solution. The rods fixated in platinum chloride were surrounded
by a yellow highly refractive ring or horse-shoe of visual purple,
whereas the outer limbs of rods fixated in Zenker’s solution
(fig. 2) looked like compact discs inside a girdle of more refrac-
tive material. The horse-shoe or ring of visual purple, stained
yellow, does not show' unless the visual purple has a fairly high

382

RTEN STENIDS.

concentration. Eyes in different stages of regeneration were
taken but it was not possible to demonstrate an intermediate
stage between the "horse-shoe” and no regeneration.

When the rods were cut along their long axis in the usual man-
ner their outer limbs stained with platinum chloride looked "in-
flated” compared with those fixated in Zenker’s solution. The
yellow colour appeared unevenly distributed giving the impres-
sion of the rods being hollow or perforated.

The cross sections certainly did give a picture of distribution
of visual purple which is in complete accordance with the views
held by Geanit and Lythgoe. On the other hand, it should
be remembered that platinum chloride penetrates relatively
slowly. This may lead to the surface being stained before the
interior of the cell which again may prevent further penetration
of the fixative into the latter. But if this were the case, it would,
be difficult to explain why the majority of the rods are surround-
ed by "horse-shoes” of stained visual purple in which the open
end of the horse-shoe often is turned towards the stained portion
of an adjacent cell. One /would then be forced to assume that
the "horse-shoes” depended on ruptures of originally complete
rings during some phase of the process.

References.

Bayliss, L. E., R. J. Lythgoe and K. Tansley, Proc. Roy. Soc.
B. 1936. 120. 95.

Geanit, R., T. Holmbeeg and M. Zewi, J, Physiol. 1938, 94. 430.
Geanit, R., A. Munsteehjelm and M. Zewi, J. Physiol. 1939. 96.
31.

Lythgoe, R, J., Brit. J. Ophthal. 1940. 24. 21.

Steen, R., v. Gkaefes Arch. Ophthal. 1905. 61. 561,

Tansley, K., Proc, Roy. Soc. B. 1933. 114. 79.

Aus den Physiologischen und Pharmakologischen Abteilungen
des Karolinischen Institutes, Stockholm.

liber den Beizmeclianisnms der Cliemorezepto-
ren iin Glomus caroticuni.^

Von

U. S. V. EULER, G. LILJESTRAITO and Y. ZOTTERMAN.

(Mit 1 Fig. im Texte.)

Heymans, Bouckaert und Dautrebande stellten 1931 fest,
dass gewisse atmungserregende Giffce vne Lobelin und Nikotin
ihro Wirkung reflektorisch. iiber das Carotissinusgebiet ausiibten.
Abnlicbes wurde spater auch fiir Azetylcbolin (Heymans u.
Mitarb. 1936) und Kaliumionen (Euler, 1938) gefunden. Die
genannten Wirkstoffe sind alle dutch ihre synaptotrope Wirkung
gekennzeichnet, und es wurde die Annahme vorgefuhxt, dass sie
prinzipiell gleichartig, d. h. erreaend an synapsahnlichen Struk-
turen, vrirken (Euler, Liljest^nd und Zotxerjun, 1939).

Wenn Aktionspotentiale vom Carotissinusnerven der Katze
abgeleitet werden, beobachtet man nach Zufuhr von Lobelin
Oder Nikotin eine starke Zunahme der Impuisfreqttenz, die sich
auf die feineren, chemische Impulse tragenden Fasern bezieht.
Die grosseren Druckimpulse bleiben dabei unbeeinflusst. In
neueren Versuchen haben vdr nachgewiesen, dass auch kleine
Mengen Azetylcholin, die am atropinisierten Tier keine Atmungs-
oder Kreislaufwirkung hatten, eine bald voriibergehende aber
starke Erhohung der Impulsfrequenz herbeifiihrten. In diesen
Versuchen wurde 6 — 10 jtig dutch die A. carotis externa dem
Sinusgebiet zugefuhrt. (Fig. 1).

Dieses Ergebnis scheint die Moglichkeit auszuschliessen, dass
es sich um eine direkte Erregung der Chemorezeptoren handeln

’ Eingegangen am 14. Dezember 1940.

26f — i01323. Acta plij/s. Scandinav. VoLJ.

384 U. S. V. EULER, G. LILJBSTRAND UND Y. ZOTTERMAN,

yv V V V V V VV VV V W V V V V"V V V V V V V Y VV V vv V V V

Fig. 1. Carotissinusnerv, Katze. Chloralose. 1 mg Atropinsnlfat pro kg intra-
venSs. A. Chemorezeptoren dnrch Uberventilation ansgeschaltet. B. Eeichlich
chemiache Impulse nach Injektion von 7 //g Azetylcholin in 0.16 ml Einger-
15snng durcli die A. carotia externa. C. Einsetzen der Azetylcholinwirkung
bei tt> 0A6 Sek. nach der Injektion bei f. Kurven von rechta nach links zu

lesen., Zeit Vso Sek.

konne, da Eeizwirkungen von Azetylcholin an sensiblen Nerven-
endigungen unwahrscheinlich und zumindestens nicht bekannt
sind, im Gegensatz zu Lobelin, Nikotin und Kaliumionen. An-
derseits besitzt Azetylcholin eine charakteristische Reizwirkung
an Synapsen verschiedener Ait, wie sie durch die Ergebnisse
iiber die Enegungsuberleitung in sympathischen Ganglien und
motorischen Endplatten bekannt ist.

Wir sehen hierin eine Stiitze fiir unsere Auffassung, dass nicht
nur Azetylcholin sondern auch die ubrigen, oben erwahnten che-
misch aktiven Stoffe durch eine sjmaptotrope Wirkung ihren
Effekt cntfalten. Das Vorhandensein von Ganglienzellen ini
Glomusgebiet steht hiermit in Einklang.

Eine Erregung der Barorezeptoren konnte mit Azetylcholin
nicht beobachtet werden.

Durch die hier mitgeteilten Befunde diirfte somit erstmalig
ein Synapsmechanismus in afferenten Nerven ausserhalb der
Sinnesorgane dargetan sein.

REIZMECHANISMDS DER CHEMOREZEPTORER.

385

Literaturverzeielinis.

Euler, U. S. v., Skand. Aick. Physiol. 1938, 80 , 94.

Euler, TJ. S. v., G. Liljestrard und Y. Zotterman, Uppsala Laka-
refor. Eorhandl. 1939, 45 , 373.

Heymans, C., J. J. Bouckaert und L. Dautrebande, Arch. int.
Pharmacod}Ti. 1931, 40 , 54.

Hey-Maxs, C., j. j. Bouckaert, S. Farber und F. J. Hsu, Arch,
int. Pharmacodyn., 1936, 54 , 129.

From The Neurophysiological Laboratory, The Caroline Institute,

Stockholm.

The Receptor of Testudo.^

KAGNAR GRANIT.

(With 1 figure in the text.)

By means of the micro-electrode technique, applied as des-
cribed in detail by Granit and Svaetichin (1939), I have recorded
with cathode ray and condenser coupled amplifier the "spikes”
of activity from single units in the cone-retina of the tortoise
{Testudo graeca). It was found to be particularly easy to isolate
a "red” receptor in this eye, in fact, the records of single units
in Fig. 3 of my recent paper in the same volume of This Journal
(Granit, 1941) are all samples of this "red” receptor. The whole
retina has bee illuminated with light from a Tutton monochro-
mator, and the energy measured necessary for a constant res-
ponse such as the absolute threshold or the cessation of “flicker”,
caused by intermittent stimuli. For energy control etc., see the
paper by Granit and Svaetichin (1939).

The curve drawn in full between the large circles in Fig. 1
shows inverse relative energies in per cent of the maximum, placed
in 0.620 fi. Behind this curve are the averages of 81 readings with
5 animals. With three of them the absolute threshold for a
single “on”-spike was studied, with the other two the cessation
of “flicker” for respectively an “on”- and an "off”-spike was
the constant index necessary for the measurements. The curve
was independent of the index used. The eyes were in different
states of adaptation but this factor also had no influence on the
distribution of sensitivity of the "red” receptor. The state of
adaptation merely determines the general level of excitability.

* Received 2 Jannaiy, 1941.

TUB ‘‘red” receptor OF TESTDDO.

387

Fig. 1.

In tliis respect the curve for the "red” receptor radically differs
from most of the visibility curves, obtained from lightadapted
eyes of frogs. Nearly all receptors •which in the photopic frog’s
eye are sensitive to long "Wave-lengths become greensensitive
during dark-adaptation and finally their -visibility curves acquire
the shape of absorption curves for -visual purple (Granit and
SvAETiCHiN, 1939) -with maxima around 0.500 u.

For comparison I have added in dotted lines between the small
filled circles the "reddest” receptor that I have found in the
light-adapted frog’s eye. Two spots in the same eye gave this
narrow, unusually stable and precise visibility curve, based on a
group of 3 — 4 very large spikes. With the electrode in the first
spot were obtained 20 readings during 60 min., from the second

388

RAQNAR 6BANIT. .

Spot 29 readings during 32 min. The maximum of the 49 averaged
observations is in 0.600 (i, a somewhat unusual result, as the
sensitivity maxima of receptors in the photopic frog’s eye rarely
go beyond 0.580 jM and, as a rule, are gathered around 0.560
The remarkable red-sensitivity of the most common type of
receptor in the cone-retina of the tortoise also dominates visibility
curves based on diffuse discharges made up of several active units.
Under such conditions the maximum in the eyes of light-adapted
frogs is between 0.550 — 0.560

Beferences.

Geanit, R., This Journal. 1941. In course of publication.

Geanit, R. and G. Svaetichin, Upsala Lakaref. Forhandl. N. F,
1939. 45. 161.

From the Chemistry Department of Karolinska Institute!, Stockholm.

All Attempt at Isolating tlie Carboliydrate
Moiety of Crystalline Ovallniniin
by Means of Cataplioresis.^

ERIK JORPES and TORSTEN THANING.

The question regarding the possible carbohydrate content of the
protein molecule has for many proteins been solved through their
crystallization. On recrystallization the sugar moiety disappears.
This has been the case with hemoglobin, edestin, pepsin, tr}q)sin,
carboxipolypeptidase, Bence Jones protein, insulin, secretin,
urease and papain. Because of the specific nature of the proteins
it seems reasonable to assume that the carbohydrate component
occuring in many of them is an impurity.

None the less, the polysaccharide of ovalbumin cannot be re-
moved by means of crystallization. The carbohydrate content
remains constant, about 1.8 per cent, in spite of 8 to 10 recry-
stallizations (Neubergkr, 1938).

We have tried, without success, another way of separating the
two components. We have submitted the crystalline ovalbumin
to cataphoresis in different buffer solutions. Contrary to ex-
pectation, the carbohydrate migrated with the protein both at
pH 3.56 and at 13.2. Of particular interest is the fact that the
two components followed each other closely even at pH 10.1 and
13.2 outside the stability region of the ovalbumin molecule
(SVEDBERG, 1939).

The crystalline ovalbumin was recrystallized three times. The
cataphoresis was performed in the usual way in the Tiselius
apparatus at 0°. Since the mobility of the ovalbumin molecule

^ Received 16 Jannary, 1941.

390

ERre JORPES AND TORSTEN THANING.

in the electric field at different pH values is well-known from
earlier work, no attention was paid to this question. The sharp
line was followed for 2 to 6 hours until the migrating solution
almost filled the upper cell. On the contents of the different cells
mikro-Kjeldahl and the quantitative Tillmans-Philippi orcin
reaction were made. No stress was laid upon a quantitative deter-
mination of the sugar in the ovalbumin. The relative sugar con-
tent of the different cells was determined and compared with
their nitrogen content.

In all experiments, amounting to 8 in all, with buffer solutions
of pH between 3.56 and 13.6, the sugar followed the nitrogen as
closely as one could expect, if allowance is made for the error
inherent in the method for the determination of the sugar.

A similar finding was made (Edman and Jorpes 1941) on sub-
mitting a highly purified, non-crystalline j6-glukosidase solution
of emulsin to cataphoresis. Here also the sugar, amounting to
5 to 6 per cent of the protein, closely followed the nitrogen at
such different pH values as 3.78 and 8.72.

The investigation was aided by a grant from the Therese and
Johan Andersson Memorial Foundation.

Beferences.

Edman, P., and E. Jorpes, Acta physiol, .scand. 1941 (In the press).
Neuberger, a., Biochem. J. 1938. 32. 1435.

SvEDBERG, Th., J. Amer. Chem. Soc. 1930. 52. 5189.

acta physiologica scandinaviga

VOL. I. SUPPLEMENTUM I.

From the Physiological Department (Head: U. v. Euler) and the
Pediatric Clinic (Head: A. Lichtenstein) of the Caroline Institute,
Stockholm, and the Physiology Institute (Head: R. Granit),

Helsingfors University.

CONTRIBUTIONS
TO THE NEUROPHYSIOLOGY
OF THE OPTIC PATHWAY

ACADEMICAL TREATISE

WHICH BY DUE TERMISSION WILL BE PUBLICLY DEFENDED AT THE CAROLINE
INSTITUTE, STOCKHOLM, ON MAY 14, 1940, AT 10 A. M.

IN THE LECTURE THEATRE OF THE PATHO-
LOGICAL DEPARTMENT

Carl Gustaf Bernhard

MED. Lie.

STOCKHOLM 1940

CONTENTS.

Page

Preface 5

Part X. On Sloiv Poiential Changes Following Light Stimuli

in the Retina and the Optic Nerve of Frog 7

latroduction 7

Components of the Retinogram and their Relation to Im-
pulse Discharge in the Optic Nerve 10

Nature and Origin of the Retinol Gomponenis 15

Slow Potentials in the Roots of the Spinal Cord .... 18

Author’s Investigations 19

Methods (Apparatus and Preparation) . . . - 20

Results 23

Introductory Experiments Demonstrating Excitation and

Inhibition in Retina and Optic Nerve 23

Responses from Different Parts of the Optic Nerve when

Light is superimposed on the Off-effect 28

Discussion and Conclusions 36

Summary 43

Part II, Observations on the Electrical Response to Light from

the Optic Tectum in Frog 45

Introduction 45

Author’s Investigations 46

Methods 47

Results 47

Discussion 49

Summary ^4

Part III. Time Correlations in Man of Eleclrophysiological and
Sensory Phenomena Following Light Stimuli 52

Introduction

Author’s Investigations 39

General Apparatus o9

CARL GUSTAF BERNHARD.

Relation between Intensity of Visual Stimulus and the
Blocking Time of Human Occipital Alpha Rhythm com-
pared with Intensity-Latency Relation of Action Poten-
tials in the Optic Nerve of Frog

Blocking Time of Human Occipital Alpha Rhythm. (Tech-
nique and Results)

Latency of Action Potentials in the Optic Nerve of Frog

(Technique and Results)

Discussion

Investigations of the Retinal Action Potential of Human Eye
with Special Reference to the Relation between Latency

of Retinogram and Blocking Time

Technique and Results

Discussion

Relation between Blocking Time and Motor Reaction Time

Technique and Results

Discussion

Relation between Blocking Time and Perception Time at

different Intensities

Technique and Results

Discussion

General Discussion and Conclusions .

Summary

Meferences

Page

80
80

84
88
90

PREFACE.

The publishing of these investigations provides me with the
opportunity of expressing my heartfelt thanks to Professor Goran
Liljestrand, who has superintended my scientific studies from
the very beginning and constantly facilitated my experimental
work by means of valuable guidance. I feel deeply grateful to
Professor Adolr Lichtenstein for the keen interest he has shown,
and the active support and generous help he has provided me
with at his clinic to enable me to carry out the experiments on
humans. I also wish to convey my sincerest thanks to Professor
Ulf von Euler for his many valuable suggestions, as well as
for all the encouragement and for the technical resources he has
so kindly placed at my disposal. I am deeply indebted to Pro-
fessor Ragnar Granit and his collaborators for the truly in-
spiring time I spent in their midst at Helsingfors. This work
is largely based on the numerous experiences that Professor
Granit ungrudgingly placed at my disposal, and I greatly
appreciate the criticism he has expressed during the course of
my experiments. I desire to express my best thanks to Dr.
Carl Rudolf Skoglund, who was ever ready to help me over
the many difficulties with which the path of the electrophysiol-
ogist is always filled. Mr. C. G. Aspenius, has given me valuable
technical help, for which I am very grateful.

Finally I should like to express my great appreciation to BEss
Dorothy Ferris for the care she has taken "with the trans-
lation.

CARL GUSTAF BERNHARD.

It has been possible for me to carry out the investigations
thanks to grants from the Foundation Therese and Johan
Andersson'’s Minne and the Nobel Foundation, as Avell as fellow-
ships from the Retzius fond of the Swedish Medical Society and
from the Lindahlsfond of the Royal Swedish Academy of Science.

Carl Gustae Bernhard.

PART I.

Oil Slow Potential Changes Following Light
Stininli in the Retina and the Optic
Jfeiwe of Frog.

Introduction.

Prom a neuropli 3 '^siological point of view the retina is an espe-
cially interesting object, since it combines an extremely differen-
tiated sense organ with a “true nervous centre” (Cajal, 1894).
The eIectroph)'siologicaI analysis of the retina and the optic
nerve which has been made dxiring recent decades (especially by
Adrian and Granit el ah), has brought to light facts of consider-
able importance for the analysis of problems concerning the phys-
iology of vision, as well as for the elucidation of certain central
nervous mechanisms.

The phenomena of the Sherrington school termed spatial
and temporal summation, synchronization and inhibition, have
their electrophj'siological correlates in the retina and have been
analysed. The distribution of the integrative processes in the
retina in relation to its morphological features gives an example
of how a functional analogy to the anatomical facts, as expressed
in Cajad’s definition, has been proved. The results, which show
that the retinal transition from scotopic to photopic vision is
largely caused by changes in the neurological state of the retina,
as well as by changes in the photochemical processes, may serve
as an example of newly discovered facts of fundamental im-
portance in the physiology of ^’ision.

The investigations made by Adrian and Matthews (1927 a
and b, 1928) of the action potentials in the retina and the optic
nerve of the eel were the first to give electrophysiological proof
of the synaptic junctions of the retina. They found in the optic
nerve the general relation between the intensity of stimulation,

CAKL GUSTAF BERNHAED.

and the frequency of discharge which had been found to exist
in other sense organs and neurones (see Adrian, 1932 a). In
the increase of frequence and the shortening of latency they
also found that an increase of the illuminated area was equi-
valent to an increase of the intensity of illumination. They made
use of the influence of this spatial effect on the latency as a mea-
surement of the interaction in the retina, and thus proved (1928)
that the latency of the impulses in the optic nerve was shorter if
4 spatially separated areas were illuminated simultaneously than if
each spot were stimulated separately. Graham’s experiments
(1932) on the Limulus eye, which is lacking synapses, point to the
fact that this spatial effect is a result of interaction transmitted
by the lateral connections of the retina (Adrian and Matthews
1927 a and b, 1928; Granit 1933). In this eye a shortening of
the latency of the optic impulses is obtained by the increase of
intensity, while a change of the area illuminated has no effect.
Adrian and Matthews (1928) showed on the spatial effect that
the lateral connections in the eye react to strychnine in the
same manner as similar connections do in the central nervous
system.

Such experimental criteria for the synaptic activity of the retina
augment the anal 3 d;ical importance of the complex slow retinal
potential following illumination which is associated with impulse
outburst in the optic nerve.

Slow 'potentials have now been found in many ganglion cell
structures elsewhere. Associated with impidse discharge in
efferent neurones, these potentials reflect in their courses different
phases in the course of central excitation.

Thus, slow ganghon potentials have been found in the optic
and abdominal ganglion of the water-beetle (Adrian, 1931, 1932 b
and 1937 a), the spinal cord (Gasser and Graham 1933; Hughes
and Gasser, 1934 a and b; Eccles and Pritchard, 1937; Hughes
et al., 1937; Matthews, 1937; Barron and Matthews, 1936 and
1938), the superior cervical ganglion (EcciiES,1935 a and b), the
ciliar ganglion (Whitteridge, 1937), ganglions of median cardiac
nerve of Limulus (Heinbecker, 1936), the inferior mesenteric
ganglion (Lloyd, 1937 and 1939) and the fifth lumbar ganglion
(Obrador and Odoritz, 1936).

The slow potential appearing in the retina following illumina-
tion, i. e. the electroretinogram, (ERG) was first observed by
Frithiof Holmgren 1865 (1880), and later, independently, by

NEUEOPHYSIOLOGY OF THE OPTIC PATHWAY. 9

Dewabe and McKendkick (1873). Gotch’s records (1903) of the
ERG were the first to give a true picture of all the different phases
of the retinogram in agreement with later and more complete
investigations. The proof that the principal course of the ERG
is the same in all vertebrate eyes (Bbucke and Gabten, 1907;
PiPEB, 1911; Habtline, 1925) is of co-ordinate importance,
the corneal electrode in the principal phase of the retinogram
being positive in relation to the electrode on the back of the bulb.
The ERG from intact animals and enucleated eyes also give in
principle corresponding results (Habtline, 1925),

Taking into consideration the complicated neuro-anatomical
structure of the generating object, it was thought at quite an early
date that the total ERG must be composed of several component
potentials of different signs and amplitude, which, in varying time
relations, give rise to the different phases of the retinal potential
(KiiHNE and Steineb, 1880; Einthoven and Jolly 1908;
PiPEB, 1911 and Kohlbausch, 1918). Einthoven, Jolly and
PiPEB studied the change in the retinogram with varying types
of stimulation, and, from the phase changes obtained, they
arrived at three different components. Einthoven and Jolly
introduced the phase signs a, h, c and d, which have since been
in general use. The important component analysis carried out
by Gbanit (1933) on decerebrated cats affords greatly increased
interest concerning the electrophysiology of the retina. Gbanit
assumed that certain phases of the retinogram should be more
sensitive than others owing to varying resistance in the source
structure. It also turned out that he could selectively produce
definite reversible phase changes in the ERG. In the course of the
phase change, Gbanit was able to identify different components
in the total effect. The different components P I, P II and P III
were fixed and analysed with reference to their relation to the
impulse discharge in the optic nerve.

Various vertebrates have served as objects for the numerous
investigations which have since been carried out by Gbanit and
his co-workers. Investigations concerning the division of the
retinogram into different components and their functional rela-
tions have shown corresponding results on the whole (frog, Gbanit
and Riddell, 1934; owl, pigeon and mouse, Gbanit, 1935; rat,
Chabpentieb, 1936). Concerning the human ERG, see pag. 70.
Individual divergencies have been ascribed to anatomical or func-
tional differences, conditions that have served to illustrate the

CARL GUSTAF BERNHARD.

principal problems from different points of view. Tbe following
representation will chiefly treat facts obtained from experiments
made on the frog.

Components of tlie Retinogram and tlieir Relation to
Impulse Discliarge in tlie Optic Nerve.

Fig. 1 shows a schematic diagram of the components in the ERG
of the dark and light adapted frog’s eye (Granit and Riddell,
1934). The initial negative a- wave is formed by the very first,
as yet uncompensated, part of the earliest appearing negative
component P III. The positive component P II is superimposed
on P III, and in its earlier course P II forms the contour of the
&-wave. The slowly rising c-wave in the dark-adapted eye after
a long period of illumination is formed by P II and the positive
P I which occurs later. P III, which lies hidden after the super-
imposition of the positive components, rises quickly to the base
line on cessation of light. P II falls slowly, and the d-wave, the
off-effect, seems to rise schematically like an interference pheno-
menon between P III and P II.

P I which is only to be found in the dark adapted eye is not
associated with any measurable change of impulse frequency in
the optic nerve (Granit 1933; Granit and Therman, 1935;
Therman, 1938). This component which is selectively enhanced
by adrenaline (Therman 1938) seems to have something to do
with the excitability in the retina, although this has not yet been
explained. When using the light adapted eye the influence of
P I is eliminated.

P II is the only component that is associated with impulse
response in the optic nerve, and by eliminating. P II the impulse
discharge in the. nerve disappears - (Granit, 1933; Therman,
1938). P II’s qualities as an excitatory component is demonstrated
by its relations to the impulse picture in the optic nerve e. g.
with variations in area, intensity and state of adaptation (e. g.
Granit, 1932 and 1933; Granit and Therman, 1935).

Thus the 6-waye formed by P II is the best electrophysiological
expression for the sensitivity of the retina. It is used as a measure-
ment of the effect from the retina when investigating the sen-
sitivity of the retina in different states of adaptation (e. g. Char-
PENTiER, 1936; Riggs, 1937; Wrede, 1937; Therman, 1938), its
relation to different wave lengths equalized with respect to energy

mV.
+ i*ot

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY.

Fig. l. Analysis of the components of frog’s retinal action potential (drawn
in full) in dark adaptation (upper set of curves) and light adaptation (lower set
of curves). Components dra'wn in broken lines. Thick black line indicates a stim-
ulus of 2 sec. duration. Explanation in text. (Granit and Eiddell, J. Phys-
iol., 1934, 81: 1.).

(Graham and Riggs, 1936; Graham, Kemp and Riggs, 1935;
Granit and Munsterhjelm, 1937; Granit and Wrede, 1937),
as well as its relation to the concentration of visual purple in
varying conditions (Granit, Munsterhjelm and Zewi, 1939).
The 6-wave in the frog’s retinogram often shows several
smaller 6-maxima when using localised electrodes on the retina,
a phenomenon that has also been observed when using the elec-
trodes in the ordinary position (see Granit, 1938). These humps
in .the 6-wave will most often be attributed to simultaneous
discharges from various fibre units (Granit and Therman,
1935). It has, however, also been possible to prove that similar
polyphasic waves at low intensities really form variations of the
retinal reaction to light (Granit and Munsterhjelm, 1937).
These small 6-maxima are independent component 6-waves and

GAEL GUSTAF BERNHABD,

are explained as being retinal equivalents to fibre potentials of
different functional kinds discovered by Haetline (e. g. 1938).
Tbe positive P II “is a sum total of component potentials -witb
different adaptability and different .time constants in general’'
(Gkanit, 1938).

Wben ebminating tbe two positive components, tbe illumination
of tbe eye gives an entirely negative response, wbicb is made up
of tbe component P III (Granit, 1933; Granit and Eiddell,
1931; Therman, 1938). Tbe negative P III in tbe frog’s eye,
isolated by means of potassium, agrees witb tbe first phase of
tbe a-wave and tbe off-effect in tbe normal retinogram, as regards
its initial fall and return to tbe base line in time relation
and rate of rise. This agreement is also to be found witb variations
of tbe stimulus (Therman, 1938; Granit and Therman, see Gra-
nit, 1938).

Fig. 1 shows tbe difference in tbe frog’s retinogram in states
of dark and light adaptation. The amplitude of tbe 6-wave di-
minishes witb increasing bght adaptation, tbe retinogram falling
towards tbe base line, while tbe off-effect shows a more rapid in-
crease to higher amplitude. The same changes witb continuing
light adaptation concerning tbe fall of tbe 6-wave and tbe increase
of tbe off-effect are to be obtained in preparations wbicb are made
to fall considerably below tbe base line after tbe 6-wave (Granit
and Eiddell, 1934:). In tbe case of partial or total elimination
of positive P II, the off-effect can take place entirely on tbe
negative side. Thus bgbt adaptation favours tbe negative compo-
nent P III.

Tbe bgbt adapted state is consequently characterized by tbe
absence of tbe component P I and a striking change of PHI,
which increases in size and shows increased rapidity wben re-
turning to tbe base Uhe, at tbe same time as the off-effect in-
creases in ampbtude. Tbe change of P III can. take place with-
out a simultaneous change in P II (the 6- wave).

As tbe off-effect was shown to be associated witb a renewed in-
tense impulse .discharge in tbe nerve (Adrian and Matthews,
1927 a and b; Granit, 1933) and P II was found to be associated
with excitability; Granit assumed that tbe reappearance of
tbe positivity of tbe off-effect witb its accompanying impulses
was to be regarded as a “post-inhibitory rebound”, inconsequence
of P III having bad an. inhibitory effect on positive P II. .

Just as P II has been shown to form a typical “excitatory

NEUROPHYSIOLOGY OF THE OPTIC PATmVAY.

potential”, the identification of P III as an “mhibitory potential”
has gained ample experimental support. As has been pointed out,
the change in the off-effect is quite opposed to that of the 6-wave
in different states of adaptation, as it increases with light adapta-
tion. This, as well as the fact that the off-effect in the negative
retinogram (P III) of the potassium-treated frog’s eye copies in
its rise the off-effect of the normal retinogram, made it probable
that the off-effect had something to do with the component P III
(Granit and Riddell, 1934; Therman, 1938; Granit, 1936 and
1938, Wright and Granit 1938),

Eintiioven and Jolly’ (1908) observed that the o-wave became
more pronounced if the retinogram was recorded immediately
upon a light stimulus. Granit and Riddell (1934) made this
observation a basis for a quantitative investigation of the participa-
tion of the components in the off-effect. The effect of a flash, su-
perimposed upon the off-effect on the light-adapted frog’s eye,
caused a strong “negative notch” in the positive off-effect. This
negativity is greatest when falling on the maximum of the off-
effect, and decreases at shorter and longer intervals between the
cessation of the light and the following flash. The strong nega-
tivity, w’hich implies a partial or total removal of the off-effect,
passes at longer intervals in the o-wave of the normal retinogram
and is accompanied by the positive 6-wave. Granit and Riddell
showed that the negativity produced by the superimposed light
stimulus is due to a selective reactivation of the negative compo-
nent P III, A flash on the maximum of the off-effect scarcely
gives any “negative notch” in the dark adapted eye, and in agree-
ment with this, the off-effect of the eye treated with potassium
shows a strongly marked “negative notch” (Therman, 1938).

By means of this method of rendering P III sslectively active
during off-effect, Granit and Therjian (1935) investigated the
relation of the nerve impulses to the negative P III, It proved
that the reactivated negativity in the retinogram corresponds to
a momentary stopping of the off-impulses of the nerve, and thus
a strong functional support was obtained for the identification
of the P III component as an inhibitory one.

Tests made by Wrede (1937) and Therman (1938) show how
the 6-wave rises during dark adaptation, while the off-effect at
first rises and then falls in amplitude. During the course of the
dark adaptation Therjlan. followed the amplitude of the off-effect
and the 6-wave at monochromatic stimuli of 0,450 n and 0.650 n

CARL GUSTAF BERNHARD.

resp. At 0.450 [j., tlie 6-wave rises gradually with dark adaptation
to maximum, while the off-effect only rises when the 6-wave has
reached its maximum, and then it again falls and disappears.
(At 0.650 {x the retinogram does not participate in these adap-
tive changes of the rod spectrum). According to Granit (1938),'
these observations might find their explanation in the fact that
the processes releasing the on- and off-effects in the dark adapted
eye divide pathways, so that the 6-wave has a refractory action
on the off-effect.

This possibility finds experimental support in important addi-
tions (Gbanit and Therman, 1937) to the investigations made
by Granit and Riddell. The part of the off-effect, not elim-
inated by the superimposed flash, can be made to disappear when
the off-effect occurs near a previous 6-wave and disappears with
the elimination of the component P II. This non-inliihitable fart of
the off-effect thus seems to belong to the component P II as does
the 6-wave. Whereas that part of the off-effect which reacts to
superimposed light , with the negative notch coinciding with the
trhpulse inhibition in the nerve, i. e. the inhibitable part, is inde-
pendent of P II. In the eye treated with potassium, the retino-
gram of which has been deprived of its component P II, the whole
off-effect IS inhibitable (Therman, 1938).

Hartline’s latest investigations (1935 and 1938) of the impulses
from single fibres in the optic nerve (bull-frog) show how different'
single fibres give different definite types of response. Some
nerve fibres respond at the onset of light with an impulse out-
burst of high frequency followed by a constant discharge of low
frequency, which ceases with the cessation of light.. Other fibres
discharge both at “on” and “off”. Finally a third group gives a
discharge only at the cessation of light,. but this is of a strong and
lasting character. The discharges of this latter group are checked
by re-illumination.

Thus, it seems (Granit and Therman, 1937), as if the non-
inhibitable part of. the off-effect due to positive P II would form
a retinal equivalent to the effect from that type of fibre which,
according to Hartline, discharges both at the onset as well as
at the cessation of light. The inhibitable part of the retinal off-
effect is, on the other hand, related to the elements, the off-
discharge of which is inhibited by. reillumination.

Veratrine, however, brings the off-effect to a state of complete
inhibitability without the 6- and d-waves undergoing any change

KEUROPHYSIOLOGY OF THE OPTIC PATHWAY. 15

of amplitude (Therman, 1938). On this ground Theeman assumes
that the distinction between the two components of the off-effect
is not absolute, while the veratrine effect indicates that the nega-
tive notch due to P III forms a significant factor for both compo-
nents.

Nature and Origin of tlie Retinal Components,

Ever since Granit’s first work in 1933, the difference between
the componente of the retinogram as well as their correlation to
opposite functional qualities have received many criteria. The
above-mentioned tests show how two potentials of opposite signs
are produced in the retina when illuminated and how these poten-
tials are associated with c.Ycitability and inhibition respectively.

Thermae’s observations on the effect of different nerve poisons
on the excitability of the retina (1938) confirm still more defin-
itely Grakit’s analysis of the components as well as the relation
of P II and P III to excitability and inhibition.

P II seems to be necessary to enable an impulse discharge to
follow in the nerve, although in certain cases (e. g. in the initial
state of potassium influence) P II can exist without impulse dis-
charge following.

P III has been shown to stand in relation to active inhibition,
which is excellently demonstrated by the association of the
retinal negativity with the interruption of existing off-impulses
in the nerve.

The opposite signs of the components in relation to the func-
tional states associated with them are of great interest from a
general neurophysiological point of view. The words positive and
negative refer to the usual manner of placing the electrodes. Tests
on an isolated retina show that in reality P II implies the negativ-
ity and P III the positivity of the receptorial side of the retina
relative to the ganglion layer (Wright and Granit, 1938; Granit
and Thermae, 1938). The reversal of the signs may be explained
by the anatomical inversion of the vertebrate retina (c. gr. Wright
and Graeit, 1938).

As regards signs and accompanying states of excitability, the
component P II generally agrees with the slow negative potential
in other ganglion structures. P II has thus been compared with
the slow negative potential in the spinal cord (Gasser and Gra-
ham, 1933) and in the superior cervical ganglion (Eccles, 1935 a

CARL GUSTAF BERNHARD.

andb). On an analogy the component P III, which is associated
with inhibition, has been generally compared with the slow posi-
tive potential in the same nervous structures (Granit, 1933;
Granit and Therman, 1935).

As a basis for extensive investigations of the relation of the
retinal excitability to different nerve poisons, Therman (1938)
proved the tenability of a close comparison of the components
P II and P III with the after-potentials in peripheral nerves and
slow ganglion potentials elsewhere.

Thermae’s observations on the reaction of P II and P III to
the different nerve poisons (potassium, calcium, veratrine, and
strychnine) do not show sufficiently, satisfactory agreement with
the results obtained on peripheral nerves (Graham, 1933; Leh-
mann, 1937; Graham and Gasser, 1931; Graham 1930) to identify
the retinal components with the after-potentials in the peripheral
nerves.

Owing to the fact that nerve poisons have hitherto been tested
to a very limited extent on slow ganglion potentials (Lehmann,
1937; Eccles, 1935 b), Therman has delayed coming to any
conclusion concerning the comparison of the retinal components
with other slow ganghon potentials.

In order to illustrate the causal relation between the retinal
potential and the impulses of the optic nerve, and in order to
investigate whether P II and P III are true opposite potentials,
Granit and BDelme (1939) tested . the influence of electrotonic
states on retinal excitability.

Retinal inside, cathode increased the on- and off-effects, both
in the retina and the nerve, while the retinal inside anode produced
a decreasing effect. Polarizing current has the same influence on
the negative component P III (obtained by treating the eye with
potassium). The fact that P II and P III seem to be similarly
influenced by a polarizing current appears to indicate, according
to the authors, that the components really are potentials of oppo-
site signs. This result makes it scarcely probable that their oppo-
site signs would be due to a different orientation of the respective
source structures in relation to the electrodes.

The experiments with polarization show how the retinal excit-
ability is influenced by electrotonic states, The authors point
out the possibility of the fact that the spread of excitability in
the retina might be of an electrotonic kind, a possibility even
expressed by Granit and Therjian (1938).

KEUROPIIYSIOLOGY OP THE OPTIC PATHWAY. 17

Proceeding from the argument that Eccles (1935 b) used as
a basis for pointing out the correspondence of the P and N
processes with cat- and anelectrotonus respectively, Therman"
(1938) illustrated the influence of poisons on the different states
of catelectrotonus and anelectrotonus in the retina, if P III were-
to be regarded as a state of anelectrotonus.

Investigations by GnAifiT (1933), Granit and Thermal
(1935) and Therman (1938) show that the ERG is not made up of
summed impulse potentials. The fact that synchronized impulse
potentials may be superimposed on the retinogram, as well as
assumptions for this have been pre\'iously discussed.

Holmgren (1880) attributed a, b, and d to the retina, and
Kuhke and Steiner (1880) showed by means of e.Yperiments on
the isolated retina the correctness of this statement. The latter
authors ascribed the whole of the retinogram to the basal ends of
the sensory cells, while Garten (1907) considered the outer part
of the receptor cells to be the source of the ERG.

An increase in the intensity of excitability causes a shortening
of the latency of the retinogram (Einthoven and Jolly, 1908;
Isiuhara, 1906; Adrian and Matthews, 1927 a; Granit, 1932)
and the optic potentials (Adrian and Matthews 1927 a and b),
A shortening of the latency is obtained also with an increase of the
illuminated area (Ishihara, 1906; Adrian and Matthews,
1927 a and b; Granit, 1933). As has already been said, Adrian’s
and ^Iatthew’s investigations show that latency shortening with
an increasing illuminated area is due to sjmaptic interaction. They
show, too, that the interval from the beginning of the c-wave to
the onset of the nerve impulses is constant. These facts seem to
imply a localization of the retinogram at a point situated centrally
to the lateral connections of the retina (Granit, 1933). Moreover,
the fact that the ERG shows typical interaction phenomenon
touching both P II and P III has also been taken as a proof of a
sjmaptic or post-synaptic localization of the retinogram (Granit
and Therman, 1935).

The grasshopper’s eye, which has two layers of neuro-sensory
cells, shows a complete retinogram with all the a, 6, c, and d
waves (Crescitelh and Jahn, 1939) resembling the ERG of the
vertebrate eye. The eye of Limulus with only one layer gives on
the other hand a monophasic retinogram (Hartline, 1928;
Hartline and Graham, 1932). Thus, one component associated

2 —

CARL GUSTAF BERNHARD.

with impulse discharge in the nerve agrees with regard to the
electrical sign with P II. This, as weU as the fact that the receptors
throughout the phylogenesis may be regarded as the same kind
of cells (Kappers, Huber and Crossby, 1936) makes it difficult,
according to Granit and Helme (1939), not to correlate at least
one of the components of the vertebrate eye with the only com-
ponent of the eye of lAmulus.

The above conclusions do not altogether support these last
mentioned observations, which seem rather to make it probable
that at least one of the components is to be assigned to the recep-
torial layer (Granit and Helme, 1939),

Granit and Eccles as weU as Granit and Helme (see Granit
and Helme, 1939) also tested the effect of antidromic volleys on
the optic nerve but received no response from the retina, which
made them conclude that the ERG, or any fraction of it, can hardly
be related to the ganglion cell layer. The authors are also under
the impression that the polarizing effect on the retinogram is not
due to the influence of polarization on this cell layer.

It has also been mentioned that Granit and Therman (1938)
and Granit and Helme (1939) suggested the possibility that the
components of the retinogram or. their fractions might show the
quality of electrotonic spreading. According to the latter authors,
a similar quality might be able to explain the somewhat contra-
dictory observations concerning the difficult problem of the origin
of retinal potential.

Slow Potentials in tlie Roots of the Spinal Cord.

In experiments on the cat and the frog, Barron and Matthews
(1938) analysed the effect in the dorsal and ventral roots of the
spinal cord at electric (via the dorsal roots) and natural stimula-
tion (brief sensory stimuli). They found both in the dorsal and
ventral roots slow potential changes (cp. Gotch and Horsley,
1891; Umrath, 1933; Eccles and Pritchard, 1937; Bonnet and
Bremer, 1938) which are not caused by summed nerve impulses
but seem to be due to spread by electrotonus of potential changes
occurring in the grey matter of the spinal cord.

These slow potentials showed, quite independently of the nature
of stimulation, always the same sign with the negativity in the
electrode placed nearest the spinal cord. The slow potential of
the dorsal roots showed signs of spatial and temporal summation

NKUROPHVSIOLOGY OP TOE OPTIC PATHNYAY.

SIS well ns occlusion. The slow potential of tlie ventral roots pre-
cedes the impulse discharge, and are checked in tlieir development
hy such stimuli as inhibit impulse discharge.

The authors give the following schematic summing-up of the
experimental results which are also discussed in relation to the
membrane theory of nervous action:

"An afferent volley in the dorsal root gives rise to depolarization
at the primary terminations in the spinal cord, observed as dorsal
root electrotonus. This sets up impulses in the internuncial neu-
rones. These impulses produce depolarization of the motor neu-
rones, which sets up an impulse discharge in the ventral roots.
Tlic depolarization of the motor ncirroncs is observed as the ventral
root cleclrotonu.s.”

Author’s investigrations.

The preceding pages give a brief account of the successful analy-
sis of the elect rophysiology of the retina which has been carried
out by Grakit and his co-workers. The points of view associated
with the division of the retinogram into components have been
presented, which components have been ascribed to different
functional states. It will be understood how interest has con-
centrated inter alia around the difficult problems concerning
the origin and kind of slow potentials appearing in the retina,
problems requiring considerable experimental work undertaken
from various angles in order to bring about a solution.

Grakit and Thermak (1938) and Grakit and Helme (1939)
stated that it is possible that the slow potentials of the retina may
show electrotonic spread. Such a proof would give a new
experimental basis for the study of the spread of excitability in
the retina.

Barrox’s and Matthews’ investigations (1938) seem to show
that the .slow potentials which they registered in the roots of the
spinal cord are the result of elcctrotonic spread of the electric
changes taking place in the grey matter of the spinal cord.

These discoveries invite to a closer analysis of the effect of
natural stimulation of the retina in the optic nerve. We here meet
■^vith complications in the combined potential picture, which is
produced by the summed more or less synchronised fibre impulses.
The off-effect, however, gives us an accessible analysis basis on

GAEL GUSTAF BERNHARD .

account of tlie off-impulse in a state of discharge being effectively
brought to a stop by superimposed hght on the off-effect (Granit
and Therman, 1935).

By making use of the technique with superimposed light on
the off-effect, the effect from the optic nerve has been inves-
tigated and compared with the potential picture of the retina ob-
tained under similar conditions.

The investigations thus made are intended to show whether
a slow potential can be identified in the optic nerve when the retina
is stimulated.

Methods.

Apparahis.

In order to register the action potentials from the retina and the
optic nerve a Loewe cathode ray tube with two beams was used (type
KSH 20/2).

In most tests the oscillograph was used in conjunction with a
4-stage push-pull coupled amplifier (C. v. Sivers, Svenska Eadio-
aktiebolaget) of a type resembling in principle the one employed
by Granit and Therman (1935) and Therman (1938). For recording
both the retinogram as well as the action potentials from the
optic nerve this amplifier was used direct-coupled. This always
gave a good and steady base line and also satisfactory proportional
amplification. The absence of drift at the base line was controlled'
in the case of each registration. Calibrations 30 and 100 V at maxi-
mum amplification, see fig. 2 A and B.

Simultaneous records from the retina and from the nerve were taken
by means of the two balanced condenser coujiled amplifiers used by
Granit and Helme (1939) in conjunction with the double cathode
ray oscillograph. Simultaneous calibrations with 100 [/, V are recorded
in fig 2 C.

Time was recorded in sec. throughout by means of a Jacquet time
marker.

“White light” was used as stimulus from a 100-watt Osram projec-
tion lamp placed at a distance of about 50 cm. from the preparation
in a shaded lamp holder with a circular liole for the emerging ray of
light. The outgoing beam passed a system of lenses and prisms by the
help of which the light was focussed and brought to fall on the pre-
pared eye. For the intensity variations Wratten neutral tint filters
were used with standardized density. In all the following illustrated
tests full intensity has been used. An iris diaphragm in the beam
allowed variations of the illuminated area.

A Compur shutter placed in the beam was used for the approximately
constant standard exposures, and this was manipulated by hand. The
Compur shutter was used in order to obtain longer gaps of darkness in
the continued illumination.. The shorter gaps of darlmess were obtained

>rEUROPIIYSIOLOGY OP THE OPTIC PATHWAY. 21

by sliding a small screen quickly tlirougli the ray of light. A water
filter was also placed in the beam.

A narrow ray was diverted by means of a mirror placed between the
shutter and the preparation. This was directed to the camera by
means of mirrors in order to give direct marking of the light stimuli on
the bromide paper.

-•Ml records were made on bromide paj)cr in a camera which was
driven by an electric motor giving no disturbances.

•» <•« m '

Fig. 2. A and 15. calibrations with tlO and 100 ,«V resp. of direct-
conplcd amplifier. C. pimnltancous calibrations with 100 ,« V of
condenser coupled amplificnt. Time in '/s "cc.

Frcparalions.

The retina and nerve preparations used were chiefly from Rana
te.mporarin, Rana csculcnta being used only in a few cases.

The preparations were carried out according to instructions given
by Gkanit and his co-workers (Graeit and RinoELL, 1931; Granit
and Therman, 1935 and 1937; Tiiermax, 1938) in order to avoid the
complicating factors to which they called attention.

Eyes intended only for the ERG were enucleated and opened care-
fully, the cornea, the iris and the outer part of the vitreous body being
removed. Prepared eyes with remaining parts of the vitreous body to
protect the retina from drying thus gave constant deflections for up-
wards of 40 mins. (cp. Thermae, 1938). A steady base line or slow drift
in the main direction of the retinogram indicates, as Thermae points
out, a good and durable preparation. Decreasing amplitude in the
ERG and opposite drift in the base line was regarded as a sign of
deterioration. Records were only taken as long as the preparation gave
constant deflections, and the experiments seldom lasted for more than
30 — iO mins.

GAEL GUSTAP BEKNHABD.

The enucleated and opened bulb was resting on cotton wool soaked
in Ringer solution and was placed on a special little ebonite holder-

The preparation was made somewhat differently for the simul-
taneous recording of the action potentials from the optic nerve with
the ERG. The eye was treated in the same manner but remained
in connection with the nerve. The nerve was dissected along the
whole of its length and was kept connected with the optic decussation.
The whole preparation thus obtained was placed on the holder. The
eye was then lifted up to rest in a thin isolated metal ring so that the
nerve, which is 5 — 6 mm. long, stood like a thin stalk and in connection
with the decussation on the brain stem. This was kept either in situ
or else a small part round the decussation of the nerve was prepared
in order to serve as a foundation for the one electrode. The preparation
was moistened with Ringer solution before each experiment, and gave
constant deflections from both retina and nerve for about 30 mins.

Thin silver chlorinated pins covered with cotton wool which was
drawn out to a thin point were used as electrodes. The ERG was taken
from the cut retinal edge and back of the bulb. To record the action
potentials from the nerve, either both the electrodes were placed on
the nerve, or else one electrode on the nerve and the other on the at-
tached part of the brain. Special variations concerning the position of
the electrodes will be described below.

During the experiment the preparation lay on the holder which
was placed on a firm stand and the electrode holder was also fixed
here. The stand was put inside an electrically shielded box, which was
connected to the earth by means of the shielding of the cables. The
preparation box was furnished with a circular hole for the incoming light.

The. experiments were carried out during all the four seasons of the
year, and in order to obtain uniform conditions in which to perform
the experiments the animals were kept in the dark and in an ordinary
room temperature for 12 hours previous to the experiment (Gkanit
and ThermAn, 1937; Thebman, 1938). Only light adapted eyes were
used, and the preparation was consequently carried out in good illu-
mination. When the preparation had been fixed on the holder, it was
left to. rest for about 5 mins., during which time it was illuminated by
the test-light, which was also used as adapting light. The- same state
of light adaptation wa,s obtained by letting the light sHne duiing the
whole experiment. When the constancy in the off-effect had been tested
a few times at regular intervals of 1 sec., the registration of the effect
followed at gaps of darkness of varying length in the adapting light,
the gaps being given at regular intervals of 1 sec. It is not necessary
to keep the eye in any particular state of adaptation, something that
Gbanit and Thebman (1937) pointed out. It is oidy important for
the adapting light to be kept within a range of intensity in which the
off-effect , does not fluctuate. As will be seen from the diagrams, the
off-effect in these tests remained constant both in the retina and
the nerve. In these conditions the off-effect of the retina was about
0.5 — O.c mV, Those preparations which showed from the first great
variations in the amplitude were not used.

NEDROPHYSIOLOGT OF THE OPTIC PATHWAY.

Results.

Ijitroductory Experiments demonstrating Excitation and Inhi-
bition in the Betina and the Optic Kerve.

Fig. 3 shows a series of tests in. which, the retinograms were
taken from a light adapted frog’s eye at different gaps of darkness
in the adapting light, i. e. the adapting light being superimposed

Kg. 3. ’Osciiiograph- .records (direottcoupled amplifier) of retinal responses of
light adapted frog’s eye. A, normal off -effect. B — G show the effect of re-illu-
mination- during' the off -effect at varying intervab after cessation of light. H,
calibration 100 p. .V. Time in Vs bm.

CAKL GUSTAF BERNHARD.

on the off-effect at different intervals. The records A, B, and C
show that the off-effect has the same amplitude through the series
of tests, which indicates that the conditions have been identical
and favourable.

The shorter the gap of darkness the greater the negativity pro-
duced by the superimposed light. The negative notch is greatest
on the maximum of the off-effect (fig. 3 D) and afterwards dimin-
ishes when the gaps of darkness are made shorter.

Kg. 4. Diagram representing data obtained from the retina in a typical experi-
ment (fig. 3) with re-illumination during the off-effect after varying gaps of dark-
ness. Curve i is the average retinal off-effect. Curve o gives the level of nega-
tive dips occurring along the retinal off-effect. The amplitudes are given in per-
centages of the maximum amplitude of the normal off-effect. Time in msec, 'with
zero time at the beginning of the off-effect.

In the earliest stage of the off-effect, the superimposed light
is not powerful enough to add anything further to the negative
component (fig. 3 F and Gr), a certain recovery from the negativity
being required to bring about this, which recovery is represented by
the off-effect (Granit and Riddell, 1934) or a part of it (Granit
and Therman, 1937),. The negative notch thus increases -with in-
creasing recovery (fig. 3 C — ^D), which occurs with great rapidity.

The following 6-wave, which is extremely shght in short gaps
of darkness, increases with increasing intervals. According to
Granit and Riddell, its amplitude increase illustrates the recov-
ery of the component P II, which thus takes place consider-
ably more slowly than that of component P III.

According to Granit and Therman (1937), however, the non-
inhibitable part of the off-effect should belong to P II, as they

NEUROPHYSIOLOGY OF THE OPTIC PATmVAY. 25

maintain that the diminished amplitude of the 5-wave at short
gaps of darkness is partly due to its falling within the refrac-
tory period of that part of the off-effect belonging to P II, and
partly to Pj II showing a distinctly slow recovery at the cessation
of the illumination.

The diagram in fig. 4 classifies the previously illustrated series
of experiments in a manner similar to that adopted by Granit
and Eiddell. The d-curve represents the average off-effect, while
a shows the level of negative dips occurring along the off-effect.
The time is given in milliseconds from the beginning of the off-
effect. In order to be able to make a comparison of the data
from the different preparations of the retina and to compare the
general course of the curves when testing the retina and the nerve,
the amplitudes are most suitably indicated in percentages of the
maximum amplitude of the off-effect.

In the figure a represents that part of the whole of the off-
effect (d) that is not removed by the negativity produced, i. e.
the non-inhibitable part, the difference between a and d forming
the inhibitable part. According to Granit and Therman, the
former should be thus connected with P II, the latter making up
the component P III.

Figs. 5 and 6 give some simultaneous records of the ERG
and the impulse discharge in the optic nerve. Figs. 5 A and 6 A
show the normal off-effect with well marked synchronised im-
pulse discharge in the nerve. Records B and 0 in both figures
indicate the obvious effect discovered by Granit and Therman
(1935) of a superimposed light stimulus on the off-effect.

The already described retinal negativity appears with re-
illumination, and parallel with this the nerve impulses are sud-
denly inhibited, thus causing a short impulse gap. Not until after
this does the fresh impulse discharge follow. As Granit and Ther-
man point out, the latency of the post-inhibitory excitation (the
distance between the on-set of the fresh light and the beginning
of the following on-impulses) increases with diminishing gaps of
darkness, see figs. 5 B, 6 B and C. This period includes both the
duration of inhibition and the latency of the excitation. It is
difficult to decide how the prolongation of the whole period is
divided between them (Granit and Therman, 1935).

The figures also illustrate the fact indicated by Adrian and
Matthews (1927 a) that the on-impulses in the optic nerve begin
during the o-wave (figs. 5 and 6 B and C). The exact relation

CARL QUSTAP BERNHARD,

between tbe appearances of the retinal excitatory component P II
and the impulse discharge in the nerve (discussed as synaptic
delay) .is equally difficult to state, the exact moment for the
appearance of the 6-wave not having been determined, as the
deepest point of the negativity represents only the time when
the positive component P II overcomes the negative component
P III (Granit, 1933; Granit and Therman, 1936).

NEUKOPHYSIOLOGY OF THE OPTIC PATHWAY.

Kg. 6. Same experiment as in fig. 5 but shorter gaps of darkness.

According to Geanit and Helme (1939), a slight retinal posi-
tivity starts at off a few milliseconds previous to the discharge in
the nerve, after which the steep off rise follows a few milhseconds
after the nerve impulse. This last mentioned fact is fuUy clear in
figs. 5 and 6. The slight early positivity which, according to the
authors, might possibly represent the commencement of the return
of the component P HI to the base line, is scarcely visible in these
records, as it, requires higher amplification.

These points show the difficulty of coming to any conclusion

CARL GUSTAF BERNHARD.

with, regard to the internal relations of the processes by means of
the relation between the latencies of the retina and the nerve
effects. The observations have been put forward because they
are clearly visible in the different illustrations. The exact latencies
will not be considered in detail in future, partly on account of the
just mentioned, reason, and partly because they are of subord-
inate importance in this connection.

Responses from Different Parts of the Optic Nerve when
Light is' superimposed on the Off-effect.

Records taken from the nerve under the same experimental
conditions, though with the use of direct-coupled amplification,
present themselves as illustrated in fig. 7. Here, as in the exper-
iment illustrated in figs. 5 and 6, the one electrode was placed
right on the optic nerve, while the other was placed on a part
which had been cut out of the brain substance, the nerve being
pinched close to the decussation. Upward deviation means the
same in this case as henceforward, viz. negativity in the electrode
nearest the retina. The effect shows a rapid rise at on and off,
and this is followed by a slower fall to the base line. Synchron-
ized fibre impulses appear on the main effect chiefly at off (cp.
Adrian and Matthews, 1927 a; Granit and Therman, 1936).
Fig. 7 D shows a normal on-effect, recorded about one second
after the previous off after the cessation of the lasting off impulses
(cp. fig. 6).

On record A the impulse inhibition makes itself apparent at the
arrow. The clearly "visible synchronized fibre impulses cease,
and the curve now shows a slight tendency to fall towards the
base line, while B and C show the effect at considerably shorter
gaps of darkness. In the two latter cases the disappearance of the
impulses is accompanied by a steep fall of the off-effect.

The inhibition of the impulses, however, never produces a fall
of the curve right down to the base line, and instead the lowest
point previous to the post-inhibitory discharge generally remains
at varying gaps of darkness as the negative dip in the retinal
records.

The following on-effect amplitude decreases -with dimimshed
gaps of darkness.

Fig. 8 illustrates the series of experiments, in the same manner,
as was used for the ERG in the diagram in fig. 4, di, giving

NEUROPHYSIOLOGY OF THE OPTIC PATHSVAY.

Fig, 7. Oscillograph records (direct-coupled amplifier) taken from the optic
nerve (see text) of light adapted frog. A— C shou' the effect of re-illumination
during the off-effcct at varj-ing intervals after cessation of light. Note the syn-
chronized wavelets in A, which disappear with re-illumination (at arrow). D,
normal on-effect. E, calibration 300 ft V. Time in sec.

the average off-effect in the nerve, and representing the
lowest point reached at different gaps of darkness.

By placing both electrodes on the nerve, a similar result is
obtained in principle. Fig. 9 illustrates an experiment where the
electrodes have been placed about 1 cm. apart, making a distance
of of the length of the nerve from the retina. The spike potentials
are here more clearly visible (in fig. 9 A synchronization) on the
curve falling towards the base line. B — -F clearly illustrate the
mhibition the impulses ceasing simultaneously with the fall of the
curve. The increase of the period of inhibition stands out espe-
cially with the diminishing gap of darkness. In fig. 9 the nerve
has been pinched close to the retina but still remains connected.
No effect is visible.

As will be seen from the illustration, records with the electrodes
in these positions show some small spike potentials previous to
the appearance of the rapid and great deflection.

CARL GUSTAP BERNHARD.

Assuming that the total effect is formed exclusively hy the
summed rapid action potentials, the curve in fig, 8 woxild
approximately illustrate an “integrative recording” and form the
frequency-time curve of the nervous discharge. The frequency
of the nerve impulses rises rapidly to an early maximum both at
“off” as -well as at “on”, after which the frequency gradually
decreases (Adrian and Matthews, 1927 a; Granit and Therman,
1935). The initial outburst of impulses is represented in the
above figures by the steep upward initial deflection. In the period

i'ig. 8. Diagram representing data obtained from the optic nerve in' a typical
experiment (fig. 7) with re-illumination during the off-effect after varying gaps
of darkness. Curve is the -average off-effect. Curve c, -gives the level of nega-
tivity occurring along the off-efifect. Curve / represents -the difference between
dj and a,. The. amplitudes are given- in percentages of the maxim'um- amplitude
of- the uninterrupted off-effect. Time in msec, with zero -time at ithe beginning

of the off-effect.

where the frequency is greatest inhibition will be most apparent
(see e, g. 7, B). When the inhibition of impulses appears, the
effect does not fall, however, to the base line, although the in-
hibition of the visible impulses is so striktog that the contour
in the period of inhibition is as free from the impulses as the un-
disturbed base line before the beginning of the off-effect (e. g.
fig. 9 C—F).

As win be seen, it is a considerable part of the total “in-
tegratively” registered off-effect of the nerve (represented by
Cl, in fig. 8) that remains when the second light is super-
imposed.

NEUROPHYSIOLOGY OF THE OPTIC PATmVAY

Fie. 9 OsciUoEraph records (direct-coupled amplifier) taken from the opUo
nerve of light adapted frog, with both electrodes on I te

A, normal off-effect. B-F show the effect of re-illummation dunng the off-rffect
at varying intervals after cessation of light. G is obtemed after pmching the nerve
between the retina and the electrodes. H, cahbration 100 u V. Time in /.sec.

Some qualities of the remaining potential have been more
closely analysed in the series of experiments in w c t e e ec
has been taken from different parts of the optic nerve.

One of the electrodes was placed 1.5 mm proximally to the
retina, after which the same electrode was placed 1.5 distally
to the decussation. The other electrode remame m . eren m
both cases, as it was apphed to the under side o t e ram,

GAEL GUSTAF BERNHARD.

Fig. 10. Oscillograph records (direct-coupled amplifier) taken from the optic
nerve of light adapted frog. “Distal recording” (see text). A, normal off-effect.
B, normal on-effect. C — G, show the effect of re-illumination during the off-
effect at varying intervals after cessation of light. H, calibration 100 /«V. Time

in '/s sec.

nerve being pinched close to the decussation. In each experiment
the active electrode was then again placed in its original position
and the effect controlled, after which registration took place when
the nerve had been pinched between the retina and the active
electrode. I’inally both electrodes were placed on the retina in
order to control the ERG.

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY.

Fig. 11. The same experiment as in fig. 10 but in records A — F “proximal re-
cording” (see text). A, normal off-effect. B. normal on-effect. C — F show the
effect of re-illumination during the off-effect at varying intervals after cessation
of light. G, “distal recording” (see text), normal off-effect. H, "distal recording
after pinching the nerve between retina and electrode. I, retinal response. Time

in Vs •■’cc.

Figs. 10 and 11 show typical records from such a series of
experiments. Recording with the active electrode near the retina
(distal recording) gives tlie result illustrated in fig. 10. As will
be seen, the preparation gave an off-effect largely resembling
that in fig. 7. A steep rising peak-like deflection changes into the
slow fall of the effect to the base line. The impulses show a syn-

ii — 401317

GAEL GUSTAP BERNHARD.

chronization to units of a frequency of about 15 pr. second. The
inhibition finds clear expression when the synchronized impulses
cease. No definite deviation towards the base-line of the general
contour of the off-effect is to be seen at the gaps of darkness when
the impulse is inhibited. At short gaps (fig. 10 E and F) the new
on-effect, on the other hand, rises from a lower level than that of
the “normal” off-effect. After a very short gap of darkness as seen
in fig. 10 G, the following on-effect rises scarcely above the base
line.

The curves recorded in fig. 11 appear entirely different, the
electrodes having been placed on the proximal part of the nerve
(proximal recording). This striking difference is caused by the
off-effect rapidly falling to the base line immediately after the in-
itial deflection (fig. 11 A). The initial effect shows, however, the
same amplitude in this case as in the above. The on-effect (cp.
fig. 10 B and 11 B) appears to undergo the same change. The
changes with increasing gaps of the inhibition of impulses and
of the superimposed on-effect are the same in both cases. These
phenomena go on, however, quite near (fig. 11 E) or on the base
line (fig. 11 C and D) in proximal recording.

The active electrode is then again moved to its first position
near the retina. Fig. 11 G shows that the effect illustrated in fig.
10 remains imchanged.

If the nerve is pinched between the retina and the active elec-
trode which retains its latter position, stimulation of the retina
gives no effect (fig. 11 H). Finally fig. 11 I shows the retina still
giving a perfect response, as regards both the positive and negative
components.

Altogether some 70 experiments have been carried out and have
covered the parts illustrated in figs. 3, 7, 9, 10 and 11. In some
cases (see below) the preparations have lasted so well that complete
series of experiments have been carried out both on the retina
and the nerve (distal as well as proximal recording) on the same
object. At each recording 10 — 20 records have been made at
varying gaps of darkness.

As regards the main principles discussed above, the experi-
ments have given uniform results. Naturally the effect from the
nerve may vary somewhat in appearance owing to the complexity
of its composition. The impulses may be more or less visible on
the main effect, the tendency to synchronization may vary, and
the return of the off-effect to the base line may show trifling varia-

NEUROPHYSIOLOGY OP THE OPTIC PATEUVAY.

Fig. 12. Oscillograph records (direct-coupled amplifier) taken from the optic
nerve of light adapted frog. A-^C, “distal recording”. "pw^mal record-

ing” (see text). D, calibration 100 /tV. H, “distdl recording after pine g
nerve between retina and electrodes. Time in sec.

tions. Eecords in fig. 12 from an experiment with distal and
proximal recording show slight differences, as regards the pheno
menon mentioned above, from the curves illustrated in figs,
and 11. The variations may be said to lie between these two
types. Fig. 12, however, illustrates the uniformity concerning the
principles discussed in the foregoing pages.

CARL GUSTAF BERNHARD.

Discussion and Conclusions.

If the effect recorded from the optic nerve via directly coupled
amplification only consisted of the summed nerve impulses, it
should, generally speaking, form the frequeney time curve of the
nervous discharge. If such were the case, the inhibition of the
impulses of the off-effect should cause the curve to fall towards
the base line, and if all the impulses were stopped, the effect
would fall entirely to the base line. There is, however, a
non-inhibitable remainder, which if projected along the course
of the off-effect usually appears as in fig. 8. This remaining
potential rises slowly and falls gradually from a prolonged
maximum 200 — 300 msec, after the beginning of the off-
effect. After 500 — 600 msec, it falls as the contour of the
off-effect.

The non-inhibitable remainder rises at proximal recording to
as much as 40 % of the maximum amplitude of the off-effect.
This can scarcely be attributed to summed nerve impulses, it
being probable that some of them remain uninfluenced by the
inhibition of re-illumination (Hartline, 1938). The course and
size of the non-inhibitable remaining potential as well as the
“impulse free” contour of the curve during the inhibition
period do not support a similar assumption (see figs. 5, 6, 9,
and 10).

The diagram in fig. 13 B shows the data from 10 series of
experiments with distal recording from the optic nerve. In five
of these cases similar series of experiments have been carried out
on the same preparation both from the proximal part of the nerve
and the retina. The latter series are illustrated in fig. 13 A
(retina) and fig. 13 C (nerve, proximal recording). In all diagrams
the amplitude values are given in percentages of the maximum
amplitude of the off-effect. A direct comparison can be made of
the amplitude values in fig. 13 B and C, because in all these detailed

Fig. 13. Diagrams representing data obtained from the retina (A) and the optic
nerve, “distal recording” (B), “proximal recording” • (C) in ten series (sec text)
of experiments with re-illumination during the off-effect at varying intervals
after cessation of light, d and d, average iminterrupted off-effect, a and a, aver-
age level of negativity occurring along the off-effect. /, difference between d,
and a,. The amplitudes are given in percentages of the maximum amplitude of
the uninterrupted off-effect. Time in msec, with zero time at the beginning' of

the off-effect.

CARL GUSTAF BERNHARD.

experiments the off-effect gave the same amplitude both in
distal and proximal recordings (about 0.4 mV). The maximal
amplitude of the retinal off was about 0.6 mV.

The total off-effect shows quite a different picture in its general
course in proximal recording (fig. 13 C) from the one we see when
the active electrode is placed closer to the retina (fig. 13 B),
although the initial maximal amplitude is of the same magnitude,
in both cases. The off-effect in proximal recording falls much more
rapidly to the base line. The non-inhibitable remaining potential
is considerably less, its amplitude scarcely reaching 50 % of the
value in distal recording. There is no reason to suppose that a
fraction of the summed nerve impulses would undergo such a
considerable change with increased distance from the retina, all
the more so as the initial maximal amplitude is identical in both
recordings.

Thus, the high degree of amplitude diminution with increased
distance from the retina furthermore strongly argues against the
non-inhibitable part of the off-effect of the nerve being made up
of summed impulses.

The remaining potential thus rather seems to be brought about
by a slow potential change.

The diminution of amplitude of this slow potential with an in-
creased distance from the retina seems to be the cause of the
off-effect falling more rapidly to the base line in distal recording.
The maximal amplitude of the remaining potential amounts to
between 40 and 60 % of that of the off-effect, i. e. about 0.2 mV.
Taken from a point 2 — 3 ham. more proximally, the remaining
potential is then at most 15 — 20 % of the maximal amplitude of
the off-effect or about 0.08 mV.

Within the period of the maximum off-effect when the impulse
discharges are most frequent, the slow potential has risen very
slightly. The total effect in this period consists mainly of summed
nerve impulses and therefore the maximal amphtude of the
off-effect in both recordings is of the same magnitude. As soon
as the remaining potential rises and the impulse frequency dim-
inishes, the and Cj curves approach each other. Later on,
500 — 600 milliseconds after the beginning of the off-effect, the
fall of the off-effect to the base line seems to be mainly determined
by the course of the slow potential. Judging by these experiments,
it seems as if the effect from the optic nerve, when illumination

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY. 39

ceases, is made up of a slow negative potential besides the summed
nerve potentials.

It must be emphasized, however, that the potential illustrated
(«i), orily forms a non-inhibitable remainder, after the elimination
of the inhibitable fractions of the off-effect. The question now
arises whether the part eliminated by the superimposed light
contams any fraction beyond the inhibitable nerve impulses; in
other words, does the superimposed hght involve a potential change
in the nerve resembling the notch of P III in the retina which is
not to be attributed to any inhibition of the impulses?

Experiments so far carried out do not support such an assump-
tion. In the majority of cases the superimposed Hght does not
ehcit any (fig. 10) or a very sHght falling tendency (fig. 7) in the off-
effect, when the inhibition occurs after the period of the maximum
of the off-effect. In other cases a sHght faUing of the curve is
visible (fig. 13), but this is only sufficiently great to attribute it
primarily to the disappearance of the impulses in the summed
effect. On the maximum of the off-effect where the impulses are
most frequent and also the inhibitory effect most pronounced, the
rapid fall of the curve must also be ascribed to the cessation of
the impulses. A similar argument best agrees with Gbanit’s
and Therman’s (1935) investigations of the impulse frequency
in the case of superimposed Hght on the off-effect.

Should such a potential of an opposite sign accompany the im-
pulse inhibition in the nerve, it might also be expected to dimi-
nish with the distance from the retina, as is the case with the non-
mhibitable remaining potential.

In fig. 13 the curve / forms the difference between and a^.
Both diagrams show fairly good agreement concerning the general
course of the /-curve, when we take into consideration registration
defects in relation to the complexity of the potentials obtained.
Such a rehable agreement as this does not support the possib-
iHty of an opposite potential when electrodes are placed differ-
ently.

The /-curve thus represents approximately the general course
of the real frequency time curve of the nerve impulses in the off-
effect (cp. Adrian and Matthews 1927 a; Granit and Therman,
1935).

On the basis of the results obtained, it will, however, be impos-
sible to exclude a slow potential of the opposite sign appearing
in the nerve with impulse inhibition. The experiments only point

aVEL GUSTAP BERNH^VRD.

to the fact that such a possible potential cannot be of such a
magnitude — as is the case in the retina — that it makes itself
visible in the method adopted and the degree of amplification
used.

In earlier records of the action potentials from the optic nerve
(e. g. Westerlund, 1912), a negative o-wave is also to be found.
In consequence of this Granit has called attention inter alia to
the difficulty existing in the analysis of the integratively registered
nerve effect, the records easily becoming distorted owing to retinal
phenomena, a circumstance also emphasized by Granit and
Therman (1935).

Discussing the origin of the slow potentials identified in the
off-effect of the nerve, we must exclude an effect from the
retina spread simply electrically. Experiments show that no effect
w^hatsoever is obtained from the nerve if it is pinched between the
retina and the adjacent electrode (see figs. 9, 11 and 12), whereas
the effect taken direct from the retina is fully pronounced. Thus
the slow negative potential in the nerve ■will neither be caused
by summed impulse potentials passing the electrode nor be due
to simple electrical spread of the retinogram.

The fact that the amplitude of the potential diminishes with
increased distance from the retina supports a retinal origin. For
propagation along the optic nerve, it seems necessary, however, for
the nerve to be intact. These circumstances seem likely to indicate
that potential changes in the retina are spread by electrotonus to
the extra retinal portion of the optic nerve.

Concerning the potential electrotonically spread to the roots
of the spinal cord, Barron and Matthe'ws (1938) found that the
amplitude was diminished to half when moving the proximal
electrode further away from the spinal cord. According to the
investigations referred to, the slow potential in the optic nerve
is reduced by more than 50 % when mo^dng the active electrode
2 — 3 mm.

On the basis of their investigations with different recordings
from the retina and the nerve, Granit and Thbraian (1938)
emphasized the possibility of an electrotonic spread of slow
potential changes in the retina. In addition to this, they found
(personal communication) in different recordings from the retina
and the nerve that when the latter was treated with cocaine,
a slow potential remained after the disappearance of the im-
pulses.

KEUROPHYSIOLOGY OF THE OPTIC PATHWAY.

The fact that the slow potential found in the nerve should be
of electrotonic character best agrees with the different results
here obtained. Other interpretations must not, however, be ex-
cluded. There is the possibility that the slow nerve potential may
possess qualities of a slow after potential. The picture obtained
with both the electrodes on the living nerve (fig. 9) might in-
dicate that the potential follows after the beginning of the im-
pulse discharge. It is, however, quite possible that the slow poten-
tial in reality begins sooner, the initial increase being slow and
followed by a steep rise after the first impulses in the nerve. The
very earliest course of the slow potentials in these experiments
cannot be determined with accuracy.

Moreover, when considering the results obtained, we must bear
in mind the interesting fact that we are not dealing with a peripheral
nerve but with a real central tract. Fasciculus opticus (Kopsch,
1935), with its typical histological qualities, which connects two
nervous centres. Seen thus, the long potential discovered is of
interest, whether it supports the electrotonic spread of slow
potential changes in the retina or is to be regarded as a slow after
potential. Arguments for impulse initiation by slow potentials
(e. g. Babron and Matthews, 1938) and the knowledge of the
relation between the after potentials and the level of excitability
(e. g. Erlanger and Gasser, 1937; Gasser, 1939) add to the
interest of the different alternatives.

The question may be asked whether the slow potential changes
appearing in the nerve can be compared with any of the specified
components of the retinogram.

As has already been stated, the off-effect of the nerve at super-
imposed light does not show such pronounced momentary falling
that it can be directly compared ■with the negative dip of P III
in the retinogram. Nor does the normal on-effect (see figs. 7,
10 and 11) show any deflection under the base hne corresponding
to the pronoxmced a-wave of the retina (see fig. 3). Further, it
is e'vident from preliminary experiments on a potassium 'treated
eye that "while the retina gives a well pronounced negative retino-
gram, no po'ten'tial change whatsoever is ob^'ained in distal re-
cording from the nerve. These facts, again, contradict a simple
electric spread of potential changes of the retina to the nerve.

Whether the nerve possesses a negative component functionally
corresponding to the retinal P III is another question far more
difficult -to answer. As has already been pointed out, such a po-

aiRL GUSTAF BEBNHARD,

tential may possibly exist, but if so, it must be of considerably
less magnitude, as it has not been visible in the experiments.

It may be of some interest to compare the slow potential of the.
optic nerve with the non-inhibitable part of the retinal off-effect
due to component P II. They have the same sign and generally
speaking show the same course along the off-effect (cp. fig. 13 A
and B). In addition to this they are bound to the same func-
tional moments, for as slow negative potential changes, they ap-
pear in connection with stimulation.

With this comparison a similar slow potential change in the
on-effect of the nerve would be expected, and that this seems to
be the case wiU be seen in fig. 10 and fig, 11. The on-effect does
not offer the same possibility for rational elimination of the nerve
impulses, however, and therefore the possible slow potentials can
not be selectively brought out. The records mentioned, however,
go to prove a change in the on-effect with increased distance from
the retina resembling that of the off-effect, which can hardly be
considered a change in the summation effect of the impulses.

These investigations show the existence of a slow negative
potential appearing in the optic nerve when the retina is stimu-
lated by natural means. With the results hitherto obtained, it
will be too early to determine whether the slow potential is a
functional analogue to a fraction of the retinogram, i, e. the non-
inhibitable off due to P II, or whether it is a slow potential
electrotonically spread from the retina. The investigations so far
carried out point rather to the latter fact.

The slow potential thus defined in the optic nerve is of functional
interest on account of its appearing with the nerve impulses when
the retina is naturally stimulated. The investigations carried out
cannot be said to give evidence for the appearance of any slow
potential change in the nerve with a sign opposite to that of the
potential associated with excitation.

It may be pointed out that slow potentials of opposite signs
can be recorded direct from the spinal cord, and are found to be
associated with excitation and inhibition respectively (e. g.
Hughes and Gasser, 1934 a and b; see also Erlanger and
Gasser, 1937), while, (Barron and Matthews, 1938) when taken
from the roots of the spinal cord, the potential changes only
show negativity. In the ventral roots Barron and Matthews
found that the development of the slow potential was checked by
stimuli which cause inhibition of the impulse, discharge. As an

KEUROPHYSIOLOGY OP THE OPTIC PATHWAY. 43

analogY to this it may be pointod out how the slow potential in
the off-effcct of the optic nerve in its earliest stage is inhibited
in its development by superimposed light (see figs. 10 and 12).

The slow negative potential in the optic nerve at off, the
general course of which these investigations are intended to show,
seems to possess in many respects qualities resembling the slow
root potentials of the spinal cord. As its qualities are further
analysed and its possible identity with certain fractions of the
retinogram are further tested, several contributions will probably
be made to the solution of the question concerning the localization
and spread of the slow potential changes of the retina.

Summary.

1) In experiments on light-adapted frog’s eye-nerve-prepara-
tions the retinogram and the action potentials of the optic nerve
taken at different distances from the retina have been examined
and have been compared at varpng gaps of darkness in the adapting
hght.

2) The experiments illustrate, in agreement with earlier in-
vestigations (Gramt and Kiddell, 193d; Granit and Therman,
1935 and 1937; Graeit and Helme, 1939), the relation between
the ncgati\’ity produced on the off-effect of the retinogram by the
superimposed light and the impulse inhibition in the optic nerve.

3) The method adopted to inhibit certain fractions of the off-
effect vdth superimposed light has been used to analyse more
closely the integratively recorded off-effect of the optic nerve.

4) The experiments indicate that at the cessation of light a
slow potential change takes place in the nerve besides the rapid
nerve impulses.

5) The slow potential docs not seem to originate from a simple
electric spread of the retinogram or to be made up of summed
impulse potentials passing the electrode.

6) The potential, which for its propagation requires the nerve
to be intact, diminishes in amplitude by more than 50 % when the
electrode placed nearest the retina is moved 2 3 mm. further
away from the retina.

7) The slow potential change appears with negativity in the
electrode situated nearest the retina.

8) The potential is checked in its development by stimuli
(superimposed light) which cause impulse inhibition.

GAEL GUSTAP BERNHARD.

9) The experiments do not give positive expression for any
slow potential change of opposite sign (positivity of the electrode
situated nearest the retina) appearing in the nerve at impulse
inhibition.

10) The possibility that the slow potential change in the nerve
should be due to the electrotonic spread of slow potential changes
in the retina is discussed. Such an assumption seems to fit in
best with the experimental results,

11) The slow potential in the off-effect of the optic nerve is
compared with the non-inhibitable part of the retinal off-effect
due to P II, They show the same signs, have similar courses and
stand in the same relation to functional moments.

PART II.

Observations on the Electrical Response to
Light from the Optic Tectum in Frog.

Introduction.

In connection ■with the above mentioned experiments con-
cerning some electrophysiological aspects of the excitatory
processes in tbe retina and the optic nerve of the frog, attempts
have been made to gather some information about the electrical
response of the optic tectum of the frog -when the retina is stimu-
lated by light.

As has already been pointed out, the \’isual pathway forms an
especially interesting structure from a neurophysiological point
of view, the optic nerve as a true central tract linking the ner-
vous centres of the retina together with higher ones. The activ-
ity of some ganglionic connections in the higher segments of
the ■visual pathway have during recent years been studied in
various quarters. The experiments have been carried out on
rabbits and cats, and have mainly concentrated on the electrical
response of the optic cortex.

The correspondence between certain structures of the cortex and
the cortical representation of different senses have been demonstrated
electrophysiologically (e. g. Davis, 1939).

The cortical electrical response to light stimuli (Fischer, 1932;
Kobxmullee, 1932) in the form of “evoked potentials” were proved
to be confined to the area of the occipital cortex which corresponds
to the structure of the area striata (e. g. Korxmuller, 1937). This
response from the optic cortex has been analysed in experiments on
cats and rabbits, on which occasions both natural stimulatira of the
retina (e. g. Bartley, 1931, 1936 a and b; Claes, 1939; Fischer,

1932 and 1934; Gerard, hlARSHALL and Saul, 1936; Kornmuller,
1932; Wang, 1934; and Wang and Lu, 1936) and electrical shocks
applied to the optic nerve were used (e. g. Bartley and Bishop,

1933 a and b; Bishop and O’Leary, 1936 and 1938).

The analysis of Bishop and his co-workers of the localized poten-

CARL GUSTAF BERNHARD.

tials of the optic cortex seems to be the most complete yet available,
and the results of other authors seem to agree with theirs in the main.

According to Bishop et al. the response consists of three interfering
series of potential waves. The first and most rapid of them consists
of at least three rapid waves of a duration similar to axon spikes. The
first of these is attributed to the axons of the afferent radiation, and
the second to cortical neurones. The third, as opposed to the others,
can be recorded below the cortex, and is considered by the authors
to represent afferent fibres from the cortex to the superior colliculus.
The second series of waves commencing at the beginning of the second
spike potential consists of a primary slow surface-positive wave, fol-
lowed by a slow surface-negative one, both having a duration of 5 — 10
msec. Finally a third series of still slower waves (duration 100 msec.)
may follow.

(With reference to the ^videspread form of cortical response con-
sisting in the abolition of the "general” potential brain rhythm see
part III.)

The response of the optic tectum of the frog to light stimuli
has not, however, been the object of investigations. A more
detailed analysis is of special interest in consequence of the elec-
trophysiological data obtained from the frog concerning the
functions of the peripheral elements and their interconnections,
the knowledge of which is essential for the study of the central
connections of the visual pathway.

The frog offers, moreover, fairly primitive aPatomical condi-
tions, and the optic tecti are well separated and easily accessible
for preparation. The optic tectum of the frog is generally com-
posed of a central gray layer and a peripheral white one, some
superficial portions of which consist of the optic afferent fibres
which have undergone total decussation. The superficial optic
bundle leads straight from the retina, while an axial bundle run-
ning more deeply passes the nuclei geniculati thalami on its
way to the tectum (Kappers, Huber, Crosby, 1936).

Author’s investigations.

The observations reported below should be regarded as a short
appendix to the investigations concerning the slow potential
changes in the optic nerve described in part I. They deal with
potential changes obtained in different recordings from the op-
tic nerve and the optic teptum in the frog. Although these in-
vestigations are of a preliminary nature, they have given results,
however, which may be of interest.

KEUnOPJIYSIOLOGY OF THE OPTIC PATHWAY.

Methods.

The apparatus was the same throughout, as described in part I
(p. 20). In all experiments the direct-coupled amplifier was
used. The same preparation Avas also used for the retina and the
optic nerve. TJie nerve was kept in connection Avith the central
parts, and the brain of the separated head Avas exposed. The
retina Avas stimulated Avith “white light”, and the effect Avas
taken from the optic nerve and the contralateral optic tectum
AA'ith the leads in different positions. The obserA'ations Avere not
limited to any particular state of adaptation. In most cases
the ])rcparation was moderately light adapted. In the experi-
ments published below, dealing AA'ith slow potential changes, the
camera Avas running sloAA'ly.

Altogether some 300 records AA'crc taken of about 20 prepa-
rations.

Results.

Fig. 14 shoAvs typical records from an experiment in which
the potential changes in the optic nerve and the contralateral
optic tectum AA'as obtained Avhen the retina Avas stimulated with
light (on-cffccts) and with the electrodes in different positions.
The diagram (retina — optic nerve — contralateral optic tectum)
on the right of cA'cry record demonstrates hoAV the electrodes were
placed in each recording. Upward deflection indicates negativity
of electrode I and positivity of electrode II.

Record 14 A bIioaa's the general configuration we know from
the obserA'ations above (fig. 9), when both electrodes are placed
on the living ner\'e. The record in fig. 14 A shows the effect more
proximally than in the experiment illustrated in fig. 9. When
the proximal electrode (II) is placed on the brain Avithout touch-
ing the optic nerve or the tectum, and the distal electrode (I) is
retained in the same position, a potential change of the shape
illustrated in 14 B is obtained. The same result is obtained when
the proximal electrode (II), instead of being placed directly on
the brain, is brought into an indifferent (i) position elsewhere on
the tissue of the head, as was in reality the case in fig. 14 B. In
both instances a more or less pronounced initial upward deflect-
ion occurs, and this in its turn is followed by a potential change
indicating slow positiAuty of the electrode (I) which remains on
the nerve.

CARL GUSTAF BERNTLVRD

Fig. 14. Oscillograph record (direct-coupled amplifier) from optic nerve and
optic tectum of frog with the electrodes (I and II) in different positions A — D
and F — G as indicated in the diagrams (retina — optic nerve — optic tectum:
intact O and killed or removed •). E, calibration with 100 u. V. Time in Vi ®cc.
For further explanation, see text.

NEUROPHYSIOLOGY OP THE OPTIC PATHWAY. 49

The picture in 14 G is obtained with electrode I still in the same
position on the nerve, and electrode II placed on the upper sur-
face of the contralateral tectum. As will be seen, the initial de-
flection is now more complicated, and is followed by an upward
wave which seems to be mritten on the long downward deflec-
tion. ^Vith electrode II in the same position on the optic tec-
tum and electrode I in an indifferent place, a monopolar record-
ing from the tectum is obtained, the outline of which is to be found
in 14 D, showing initial downward deflections followed by a
slower wave upwards.

A\Tien the leads are replaced in the position as shown in fig.
14 B and any possible effect coming from the optic tectum is
eliminated, the record shows a picture as illustrated in 14 F,
whether the nerve is pinched in the mcinity of the decussation,
or the whole of the optic tectum is killed (by means of prolonged
potassium action) or removed from above. The configuration
of record 14 F resembles, as was expected, the picture obtained
in “proximal recording” in fig. 11.

Finally 14 G shows a record obtained after the electrodes had
again been placed in their original positions, as seen in fig. 14 A.

All the records shown here demonstrate typical potential
changes obtained at onset of light. Concerning the general shape
of the off-effect, consistent results were obtained with the on-
effect.

Discussion.

As is obvious from the records, the optic tectum of the frog
gives no spontaneous rhjdihmic potentials of a sufficiently great
magnitude to be visible with the amplification used. Even with
intensified amplification, it is difficult to identify a spontaneous
rhythm, which can definitely be distinguished from small acciden-
tal disturbances on the base line (cp. Libet and Gerard, 1939).

When the retina is stimulated by light, a series of potential
waves occurs in the optic tectum, the typical appearance of
which is illustrated with monopolar recording in fig. 14 D. The
potential response begins with a negative wave, with 2 3 crests

or else is divided into 2 — 3 waves, on which small rapid po-
tential oscillations seem to be superimposed. The amplitude of
this initial negathdty varies between 0.2 and 0.5 mV, and its
duration is approximately 100 msec.

401317

CAKL GOSTAI* BERNHAKD.

The negative potential complex is followed by a slow positive
potential of varying duration with an amplitude which is
generally somewhat lower than that of the negative potential.

No subsequent rhythmical potentials caused by light have
been found.

The first visible upward deflection in record 14 B is probably
to be regarded as the initial part of the potential of the summed
nerve impulses (cp. fig. 14 A). This initial negativity in the
electrode placed on the nerve is followed by a slow opposite
deviation which will be referred to the electrode I being put
on the nerve, it being of no consequence where electrode II
is placed. It is also visible in the recording from the nerve (I)
and the lobe (II, 14 C). The potential events from the nerve
here merge with the potential changes from the electrode on the
optic tectum. The result of this is a disintegration of the initial
potential and a change of the following slow deflection, which
can be explained by the slow positive potential of the tectum
(14 D) being superimposed. It is an interesting fact that the slow
positivity recorded from the nerve does not appear until after
the initial nerve impulses (14 B), and also that it completely dis-
appears when the optic tectum is killed or removed (cp. 14 B
and 14 F).

In consequence of the different records discussed, it seems
almost as though a slow potential change occurs in the. nerve
when this is cormected to the intact tectum. In what manner
the optic tectum contributes to its origin will not be discussed
at the present stage. Both electric and physiological spread of
the potential is possible.

Comparative experiments by placing the electrode I proxim-
ally and distally on the nerve show that the negative deflection
attributed to the summed nerve impulses at distal recording
has time to reach a more complete development before the posi-
tive deflection occurs, at the same time as the amplitude of the
latter is lower. These experiments are of subordinate interest,
however, as the potential events in the optic nerve discussed in
Part I must be taken into consideration. It is peculiar, that the
records 14 A and 14 G (before and after the elimiuation of the
optic tectum) do not show any great difference, a circumstance
that must be a -warning against hasty conclusions concerning the
interpretation of this slow potential change.

These preliminary and purely descriptive results at present do

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY.

not allow of any more detailed interpretation of the origin and
nature of the frog's tectum response to light stimulus.

It is probable that one part of the initial negative deflections,
obtained from the optic tectum (14 D), like the initial spike
potentials in the records of Bishop et al., are to be attributed to
the afferent fibres. Possibly some part of the negative complex
is due to intertectal neurones, either as fibre potentials (cp. optic
cortex, Bishop and O’Leary, 1938) or else as negative after
potentials (cp. spinal cord, Hughes and Gasser, 1934 a and b).

As an analogy, the following slow positive tectal wave might
be compared with the positive intermediary potential led from
the dorsum of the spinal cord, wdiich Gasser ct al. attributes to
the internuncial neurones being a positive after potential (Hughes
and Gasser, 1934 a and b).

According to Barron and JIattheivs (1938), the intermediary
positive potential led from the dorsal surface of the cord is the
same as the slow negative potential of the dorsal roots, and
arises, according to the authors, at the terminations of the dorsal
root fibres themselves.

Whether there is any direct connection between the positive
potential recorded from the surface of the tectum and the slow
potential change in the optic nerve must remain an open question
until the latter, which is indicated in fig. 14 B and C, has been
more closely investigated.

Summary.

1. The electrical response of the frog’s optic tectum, when the
contralateral retina is stimulated with light, has been studied
on moderately light adapted isolated head preparations.

2. The monopolar record of the tectal response, the general
configuration of which is the same at “on” and “off”, consists
of an initial negative wave with two crests or is divided into
two waves, followed by a more slow positive potential wave.

3. Recordings with the electrodes in different positions on the
tectum and the optic nerve suggest a slow positive potential in
the optic nerve following stimulation of the retina. This slow
potential change seems to disappear when the optic tectum is
kiiled or removed, or the nerve is pinched close to the decussation.

PART III;

Time Correlations in Man of
Electrophysiological and Sensory Phenomena
Following Light Stimuli.

Introduction.

Besides the "evoked potentials” which are localized to the
optic cortex (see Part II), light stimulus produces a widespread
form of electrical response from the cortex, which, finds expression
in an obvious change in' the cortical rhythmic potentials.

The human brain potentials first registered by Berger (1929)
from intact skull show typical synchronization to certain definite
frequency units. Among these dominates the one defined by
Berger as the alfha rhythm of about 10 waves per sec. Distinct
limitation and predominance (Grass and Gibbs, 1938), regul-
arity (Rohracher, 1938) and constant rhythmicity (Jasper and
Andrews, 1938; Bernhard and Skoglund, 1939) are qualities
of alpha rhythm which are visible in nearly all of the investiga-
tions made of human brain potentials. Alpha rh 3 d;hm, which
shows a certain periodicity (see fig. 15), is identified from almost
any part of the skull (Berger see 1938, Adrian and Yamagiwa,
1935; Loomis et al. 1938; Rubin, 1938). It shows the greatest
amplitude (Adrian and Matthews, 1934 b; Adrian and Yama-
GiWA, 1935) and is most amply represented (Rubin, 1938) when
taken from the occipital region.

It has not been considered appropriate here to relate the whole
extensive literature on human brain potentials under physiological
and pathological conditions, some of which are of a purely de-
scriptive nature and in this case of subordinate interest. The

NEUROPHXSIOLOGY OF THE OPTIC PATHWAY. 53

literature has been exhaustively treated by several authors (e. g.
by Bertrand, Delay and Guillain, 1939; Davis, 1936 and 1939).
It has likewise been considered irrelevant to go closely into the
different opinions concerning the mechanism of autonomous
rhythmic actmty of central neurones represented partly by e. g.
Kdbie (1930) and Lorente db No (1934, 1935, 1938), and partly
by e. g. Gerard (1936, 1937; Dubner and Gerard, 1939; Libet
and Gerard, 1939).

At existing sequences of rhythmic waves, a light stimulus
causes a disappearance of the regular rhythm. The effect, which
in experiments on human beings is visible as a more or less
complete abolition of alpha rhythm, has also been proved in the
case of animals, when the experimental conditions have been such
as to favour the rise of rhjdhmic brain potentials comparable with
the alpha rhythm of human beings (Ectors 1935 and 1936,
rabbits). The phenomenon is easily proved from intact human
skull and has been verified by most scientists in this field.

. Diverging opinions as to the interpretation of the abobtion
phenomenon are to be attributed to various theories concerning
the origin of the human brain potentials, which is at yet vague.

Experiments on animals have shown the dependence of the
cortical activity upon the deeper cortical layers (e. g. Dusser
de Barenne and Me Cdlloch, 1936 and 1938) and the sub-
cortical connections (Bresier, 1937 and 1938 a and b), as well
as that slow rhythmic waves can originate in the white matter
of cortex (Sjostrand, 1937).

Berger, who was the first to discover the phenomenon, gave
a generally formulated explanation founded on the opinion that
every part of the cortex in activity is to be regarded as the source
of alpha rhythm (1934, 1935, 1936). According to Berger the
activity in certain cortical areas caused by onset of light would
produce widespread inhibition in adjacent cortical areas resulting
in the suppression of the alpha waves.

In experiments on cortical response, Jasper (Jasper and
Andrews, 1936; Jasper, 1936) tried to find expression for the
changes in excitability comparable with observations from peri-
pheral nerves (see e. g. Erlanger and Gasser, 1937). With the
general principles for peripheral nerves as a basis, Jasper en-
deavours to explain, inter alia, the cortical reaction to hght. He
maintains that the records of the human brain potentials show
that the first period of the abobtion {alpha hloching) is associated

KEUROPnYSIOI.OGY OF THE OPTIC PATHWAY. 55

vdth .slow po.sitivitY, wliicli may be followed by slow negativity,
and that a certain rbytlimical activity may again be \dsible on
tbe latter. The author compares these suggested slow potential
clmngc.s with the negative and positive after-potentials, which,
associated with supernormal and subnormal e.vcitability, have
been proved to exist in e. g. the peripheral nerves, the spinal
coni (.«ee Kni.AXGF.n and Ga.'jseu, 1937) and the superior cervical
ganglion (EcciiE.s, 1935 a and b).

At the cessation of illuniin.ation the alpha waves gradually
return after vnrj'ing intervals of time (see c. g, Adrian el al. 1934 b
and 1935; Jasper and CRmcK.sriANK, 1937; Cruickshank, 1937)
and then .show an increase of the frequency (Cruickshank,
1937). which, aw'onling to Ja.sper (1936 b), may bo as much as
io-no

Jasper compares this phenomenon with the changes in the
injury discharge of nerve fibres occurring during the response to
an induction shock (Gasser, 1935), .showing dimini.shed activity
during the po.sitivitY, followed by supernormal activity. Suggesting
that the cortical potentials are central fibre potentials of slow
time char.icteristics. Ja-SPER explains these cortical reactions
without re.'^orting to other mechanisms than those which "can
be domon.strntcd in the nonmeduliated peripheral axon”.

Adrian and 3rATTHEW.s (19.34 b) as well as Adrian and
Yamagiwa (1935) carried out extensive scries of experiments to
prove Berger’s observations concerning the characteristics of the
liurnnn brain potentials. In experiments with the electrodes in
variou.s places, they found that the amplitude of the alpha waves
i.s greatest over the occipital region, and decreases the further
away tbe cleclTodes get. Their experiments with intermittent
light show how the potential waves in certain limits can be made
to follow the frequency of the flickering light (cp. later observa-
tions of Loomi.s ct al, 1936; .Tasper 1936; Dubup and Fessard,
1935; Gomeman, Secad and Segams, 1938). The rh}d;bmic poten-
tials then show a distribution over the bead with the frequency
of the flicker instead of the usual 10 pr. sec. rhythm. The alpha
blocking was also seen to be most satisfactorily pronounced over
the occipital region, The authors point out that the most im-
portant factor for the appearance of the alpha rhythm is the
absence of pattern vision, and as Berger points out, rest in
darkne.ss gives the best conditions for the registration of tbe alpha
frequency.

C^VRL GUSTAF BERNHAED.

According to Adrian et ai!., the experiments point to the fact
that alpha rhythm arises in consequence of co-operating spontane-
ous discharges from a large group of cortical units with occipital
localization. When these neurones are free from disturbing stimuh,
they beat in unison at their natural period. The light stimulus
produces a non-uniform excitation causing desynchronization.
According to Adrian et al., the alpha blocking thus gives expres-
sion for asynchronized activity in certain cortical areas mainly
concerned with vision.

Adrian’s investigations (1937 a) of the synchronized reactions
in the optic ganghon of the water-beetle {Dytiscus marginalis) give
an interesting analogue to the alpha blocking reaction in the
cerebral cortex. . Fresh preparations of the Dytiscm optic ganglion
give no spontaneous rhythm in darkness, while stimulation of the
receptors with strong light causes potential oscillations in the
ganghon (“bright rhythm”). The regular potential changes com-
mence with a frequency of 20 — iO per sec. slowly to approach
16 — 25 per sec. at continued illumination. If the preparation is
left ■ for some hours, a strong spontaneous rhythm in darkness
of 7 — 10 waves per sec. (“dark rhythm”) gradually starts. . In this
state bright illumination gives blocking of the dark rhythm. In
the former case weaker intensities of stimulus give an irregular
discharge, while in the latter they produce an incomplete blocking
of the dark rhythm. The potential waves of the ganglion, which
are in both cases associated with corresponding groups of impulse
discharge in the post-ganglionic nerve must, according to Adrian,
be interpreted as synchronized activity in a large number of
units.

The definite potential rhythms represent states favouring the
synchronization of the activity of the separate elements. As the
experiments show that the neurones respond over a wide fre-
quency, Adrian maintains that a fixed frequency response in the
neurones contributing to the waves cannot be assumed. The
main assumption for synchronization seems to be that the
generating units possess the same degree of excitation, so that they
beat with the same frequency. The greatest probability for- this
is to be found at the extreme ends of the visual scale, i. e. at bright
illumination and complete darkness. Bright stimulation brings
about discharges of maximal frequency, while in darkness the
neurones beat at a minimal rate ("resting discharge”). Both
states favour the synchronization of the activity. The fact that

NEUROProSlOLOGY OP THE OPTIC PATHWAY. 57

the dark rhythm only occurs in the ganglion after it has suffered
injury, on account of its being left in a reclining position —
would, according to Adrian, be due to the fact that the slight
injury produces a breakdown in the normal insulation. For the
sake of comparison, Adrian points out similar effects of injury
in the phrenic nerve (Adrian, 1930) and in the lateral line nerve
(Hoagland, 1933). Thus according to Adrian, the blocking of
the rhythmic potentials seems to arise when the stimulated neur-
ones respond to different frequencies causing a desynchronized
activity.

The synchronized effect in the optic nerve of the eel (Adrian
and Matthews, 1928) and the frog (Granit and Therman, 1935)
must be remembered in this connection. The synchronization of
the fibre impulses are best seen in the off-effect after strong stimuli
(Granit and Therman, 1935; see also figs. 5, 6, 7, 9 and 10) but
also in the on-effect in connection with stimulation of a large
part of the retina.

The tendency of Dytiscus preparations to give definite potential
rh 3 rthms, which in certain circumstances can be made to dis-
appear without noticeable frequency changes, offer an amazing
resemblance to the alpha blocking in the cerebral cortex.

The comparison between the potential changes in the Dyiiscus
optic ganglion and the human brain potentials (Adrian, 1937 a)
implies, according to Adrian, that a fixed frequency of the cortical
neurones need not necessarily be pre-supposed for the genesis of
the alpha waves (cp. Adrian and Matthews, 1934: a). When the
cortical neurones are free from external stimuli, they beat at
their miniTniiTn rate, causing the synchronization of the constant
alpha rhythm. The breakdown of this takes place in consequence
of a non-uniform excitation of the cortical units, which is prim-
arily caused by light stimuli.

The above formulated opinion of Adrian e^ al. concerning the
alpha blocking, formed after experiments on human beings, seems
to find a certain support in the results from comparable experi-
ments on simpler objects. Their explanation is also supported
by observations made in various quarters concerning the rhythmic
discharges in the nervous system (see e. g. Adrian, 1932 a, 1937 b).

The analysis of the alpha blocking phenomenon, however, is
obscure owing to several complicating factors.

On the basis of their experiments with simultaneous recording
from several points on the head, Adrian and Yamagiwa (1935)

CABL GUSTAF B3ERNH.\KD.

favour tvo occipitall 7 localized main foci for tlie alpha rhythm,
which they attributed to the two occipital lobes. The results, too,
seem to indicate that these foci cha-nge positions within certain
boimdaries. In this connection they point out that the alpha
rhythm from different parts of the head shows obvious differences
in phase between the different regions. It has subsequently been
asserted from different quarters that these phase differences are
of such a nature that there is reason to suppose that further foci
for the origin of the alpha frequency, are localized to other regions
as well (Jasper and Andrews 1938; Lindsley, 1938 b; .Janzen
and Kornmuller, 1939). The alpha rhythm with its characteristic
qualities is, however, always identified as the dominating rhythm
in the different regions (see also Davis and Davis, 1936; Loomis
et ah, 1938: Rubin, 1938).

The blocking reaction, as has been proved from several quarters,
is not peculiar to hght stimuli, but.is also caused by other sensory
stimuli (e. g. tactile, acoustic), which has been proved both in
experiments on animals (e. g, EcTORSj 1935; Rempeb .and Gibbs,
1936) as well as on human beings (Adrian and Matthews,
1931 b; , Dubup and Eessard, 1935; Jasper, 1936 b; Jasper and
Cruickshank, 1937; Rohracher, 1937; Loomis et al. 1938;
Travis and Barber, 1938),

As was pointed out by Adrian and Matthews (1931 b), the
alpha blocking seems, however, to be most intimately associated
with light stimuli, which produce the most constant and best
pronounced effects. It seems to be characteristic that just as
the typical alpha frequency is identified from different regions,
the blocking of the alpha rhythm after sensory stimuli seems to
take place simultaneously over the whole cortex. Looms et al.
(1938), who by stating the time in 1/100 sec. proved that such
seems to be the case, sum up their observations with the following
words: “With respect to disturbance potentials appearing on the
surface of the skull, the cortex acts as a whole” (see fig. 15),

"While the blocking of the alpha rhythm after light stimuli
can be recorded with the greatest facility from intact human skull,
it seems doubtful whether a local effect originating from the optic
cortex (cp. part II) can really be identified under these experi-
mental conditions.

In a few cases (Jasper and Cruickshank, 1937; Cruickshank,
1937; Jasper and Andrews, 1938), a similar local on-effect has

neurophysiology op the optic pathivay, 59

"been suggested, whicli seems to offer resemblances to tbe on-effect
demonstrated on animals (see Part II). According to Jasper, this
on-effect is best seen if tbe light stimulus is given in a period
vrben the alpha rhythm is lacking, for example, in the blocked
period after a previous light stimulus. The amplitude and latency
of the local on-effect, according to Cruickshank, would be a
function of the stimulus intensity, and she gives the value of the
latter as O.OG seconds for higher light intensities.

Some authors have pointed out, though without any actual
experimental basis, that the flickering potentials proved by
Adrian and JIatthews might possibly be regarded as evoked
potentials (Loomis et al. 1938).

The prolongation of the latency for the blocking of the alpha
rhji;hm (blocking time according to Jasper) with decreasing
light intensities was first observed by Durup and Pessard (1935,
1936). They show in a series of experiments on a person with
marked blocking of the pronounced alpha rh}dhm a continual
prolongation of the blocking time with decreasing intensities.
Cruickshank’s (1937) experiments on the blocking time as a
function of the stimulus intensity show in principle the same
result. By plotting blocking time against log brightness in some
cases, however, she finds a diphasic course in the curve thus
obtained. The diphasic curve is compared with the dark adapta-
tion curve (Haig and Wald) presented by Hecht (1934), which
■would imply that the two phases would reflect rod- and cone-
functions respectively.

Author’s Investigations.

The investigations to be recorded below thus centre around
some time characteristics of the blocking reaction of the human
occipital cortex, showing the time relation between the retinal
and cortical electrical response and the relationship between these
peripheral and central electrical responses to light and to the per-
ception of light.

General Apparatus.

Por registration a Loewe cathode ray tube (cp. page 20) with
two beams was used.

CABL GUSTAT BEBNHAKD.

Amplifiers:, For recording, hjiman brain potentials the oscillograph
was used in conjunction with two different amplifiers. The i-stage-
push-pull coupled amplifier (page 20) with a resistance-capacity-coupling
at the input (calibration see fig. 17 A) and a 4-stage condenser coupled
amplifier with 2 push-pull initial stages were used alternatively. The.
amplifiers were generally used with shunted inter-stage-coupling cap-
acities in order to eliminate frequencies of above 20 p/s (calibration
fig. 17 B), In order to see that the different condenser-couplings did
not cause any errors when determining the latency for the alpha
blocking, the effect thus obtained was controlled by comparing with
that obtained by direct coupled amplification.

Kg. 17. Calibrations with 30 'a V of direct-coupled amplifier when- using re-
sistance-capacity coupling at the input (A) and shunted inter-stage-coupling
capacities (B). Time in 1/6 sec.

For recording the human retinogram the 4-stage push-pull coupled
amplifier was used direct-coupled (calibration fig. 2 A and B). At simul-.
taneous registration of the retinogram and brain potentials this was
coupled to the one beam of the oscillograph, while the condenser
coupled amplifier was connected to the other beam used for the brain
potentials.

In the experiments in which muscle action potentials from biceps
were recorded, the condenser coupled amplifier was used.

Electrodes: Ag — ^AgGl electrodes were used in all instances. Mono-
polar recording of the brain potentials was used, the active electrode
being placed over the occipital region in the midline and the indifferent
electrode on the right processus mastoideus (Adrian and Matthews,
1934 b; Rtxbin, 1938). The electrodes used were of the construction
indicated by Bernhard' and Skoglund (1939), comfortably fixed by
rubber bands.

The same electrodes were used for recording the muscle action
potentials. They were then placed over the biceps at a distance
of 5 cm. from each other (cp. Altenburger, 1937), For the record-
ing of the retinogram, the one electrode consisting of a chlorinated
silver thread in communication with a wick soaked in physiological
NaCl solution was placed in the anaesthetized (novocaine) conjunctival
sac (cp. Cooper, Creed and Granit, 1933). The other electrode

NEUROPHYSIOLOGY OP THE OPTIC PATHWAY.

Fig. 18. Diagram of apparatus used for
light stimuli. Explanation see text.

described above was placed on the
right processus mastoideus.

The experiments were carried out
in a sound proof, ehctrically shielded
chamber for the object and an adja-
cent room for the apparatus and
the experimenter. The shieldings of
the experiment chamber, the cables
and the experiment apparatus were
earthed to a common point.

During the experiments the ob-
, _ ject rested comfortably in the dark,

undisturbed' by accidental stimuU, while the recording took place m
aCoent Toom from which the light were ako tebutA

Under these circumstances speoia i*?^;,tion both

schematically in fig. 18 A and shaded lamp holder with a

200 whtt Osram projection lamp (a) m a ^ P f jg^ses,

ctoular hole for l-c Vsm of^gh^ Pa-8 „

object._ The Ught was there was a special arrangement

sage into the chamber. In the locus ^ ^^pp^ratus is

placed for the onset nnd the mtetrupfon >‘8“ „
illustrated schematically m fig. 18 -taree

CAKIi GUSTAF BERNHARD,

(1, 2 and 3) made to turn roimd an axis (4) are held in the position given
in the diagram by means of a block arrangement (5). When this is
pushed one step in the direction of the arrow by a distance controller
operated by hand, the screen 1 falls. The light then passes through a
filter (6) in the second screen. When the block is pushed another
step, screen 2 falls and then the light is let in with its full intensity.
A last push of the block brings about the interruption of the light when
the third screen falls to take up the position given in the diagram for
1 and 2. The screens are checked in their fall to the desired position
by special brakes. The apparatus makes it possible to superimpose
stronger light on a former weaker light stimulus in a simple manner.
The intensity of the two stimuli can be varied by putting in filters
with different densities, both in screen 2 and in a holder placed just
behind screen 3. For the distribution of only one light stimulus,
screen 2 is dropped to a position where it does not function, and light
is put on and off by means of screens 1 and 3, as described above. A
little mirror fixed to the edge of the first screen throws a small beam
when the screen falls, which via other mirrors (see fig. 18 A, e) produces
marks on the bromide paper to indicate the onset of the first light
stimulus on the records. In the same manner a little mirror is inserted
behind the apparatus to divert a small beam of light in order to mark
the stronger superimposed light stimulus.

Wratten neutral tint filters were used for the variation of the
intensity of the stimuli. In the experiments published below the
stimulating disc was of 40 cm. in diameter at a distance of 60 cm. from
the eye of the oBfect, making 37° of the visual angle and without any
filter of about 10 ml. brightness.

This lighfintensity will be referred to in the text as 1, other inten-
sities as fractions of 1.

Time markings were obtained from a 50 (c./sec.) generator coupled
either to the one beam or to a neon lamp with a fine slit.

The camera was driven by a motor giving no disturbances.

Relation between Intensity of Visual Stimulus and
Blocking Time of Human Occipital Alpha Rhythm
Compared with Intensity-Latency Relation of
Action Potentials in the Optic Nerve of the Frog.

Blocking Time of the Human Occipital Alpha Rhsd^hm.

Technique.

In each particular case considerable variations occur concerning
the predominance of the alpha rhythm in the frequency spect-
rum obtained. By means of automatic Fourier Transform,
Grass and Gibbs (1938) clearly demonstrated the predominance

NEUROPHYSIOLOGY OP THE OPTIC PATHWAY’. 63

of the alpha rhythm, even in cases where it was difficult to identify
on direct inspection.

In the following only such cases liave been selected which
show clearly pronounced alpha rhythm, so that the blocking
caused by the light stimulus is distinctly marked.

In order to obtain a constant sensitivity in the retina, suitable
for the intensity .scale tested, the subject was dark adapted for
about 30 mins, before the experiment took place. After this time
the decrease of the intensity threshold in the retina approaches
asymptotically final value (Koiilrausch, 1931).

The subject lay comfortably in the dark. Attention was paid to
avoid drowsiness and sleep, conditions which change the brain
potentials and the cortical reactions (see Blake and Gerari),
1937; Davis d al. 1938, 1939; Loomis cl. at. 1936, 1937, 1938).
The experiments were carried ont on one eye. At each experiment
several intensities w'crc tested and each light stimulus lasted for
about 1 sec. Between each of these there was an interval of at
least one minute, partly to give anj’’ possible after-image time
to disappear (Jasper and Cruickshank, 1937; Travis and Hall,
1938), and partly to give the alpha rh 5 ''thm time to return to the
normal frequency value after the rise that usually appears after
the blocked period (Jasper, I93G and Cruickshank, 1937).

In order to obtain uniformity and a correct judgement of the
experiments, attempts liave been made to put in the light stimuli
in periods of well pronounced alpha rhythm.

Results.

Experiments carried out under these conditions show a
striking blocking of the alpha rhythm after the light stimulus
(sec figs. 16 and 19). The blocking usually takes place with a
sudden interruption in the regular rhythm (see fig. 19
The last alpha wave sometimes shows somewhat diminished
amplitude before the blocking, a condition which is generally
more often \dsible after low light intensities (fig. 19 E). The
phenomenon is of varying nature, and a possible temporary
diminution of tlie amplitude quite independent of the light sti-
mulus cannot be excluded (cp. fig. 15). This fact sometimes makes
it difficult to form an exact judgement of the blocking time,

Eig. 19 shows typical records from a series of experiments m
whicli six different light intensities have been tested, and illustrates
the prolongation of the blocking time with decreasing intensity.

CARL GUSTAF BERNHARD.

Fig. 19. Oscillograph records showing the blocking of the occipital alpha rhythm
to light at intensities 1 (A), 1/10 (B), 1/100 (C), 1/1,000 (D), 1/10,000 (E) and
1/100,000 (F). Time in 1/50 sec.

Tatle 1 clearly shows the good agreement in blocking time at
different intensities when the experiments are carried out under
the conditions stated above. The values are given according to
the reading in 1/100 seconds. It should be pointed out that the
variations, besides being caused by the possible biological varia-
tions, are also largely brought about by the variations in the
evaluation of the readings, as stated above. As is seen, the
deviation from the average remains within reasonable limits.

In fig. 20 the average values for the blocking time at different
light intensities obtained in experiments on five different human
subjects are plotted against log brightness. The different persons
are represented in the figure by the five different tj’pes of
circles marking the individual average values.

As is evident from the diagram, the values for the blocking
time at the different intensities for the different persons are of
the same magnitude. The blocking time is seen to increase con-
tinually with decreasing intensities of stimuli. The curve graphic-

IfEUROPHYSIOLOGY OP THE OPTIC PATHAVAY.
Table 1.

CARL GUSTAF BERNHARD.

ally obtained slxo-\vs tlie general relation between the blocking
time and the log brightness within the range tested of the visual
scale, which lies between the value for the visual threshold and the
intensities at Avhich the blocking time asymptotically approaches a
minimum value.

Latency of Action Potentials in the Optic Nerve of Prog.

Technique.

The action potentials were recorded by means of the condenser
coupled amplifier described by Granit and Svaetichin (1939)
with short time constant and balanced input stage. In addition
to this the registration apparatus described on page 20 was
used.

Both “white light” and monochromatic light was used. The
spectral light was obtained from a Tutton monochromator
described by Granit and Munsterhjelm (1937), which was used
according to Granit’s and Svaetichin’s (1939) instructions.
The same technical arrangements with regard to “white light”,
preparation etc. were made as previously described on page
20 .

The experiments were made on preparations from dark-adapted
frogs, which were kept in the dark and in room temperature for
at least 12 hrs. before the experiment was commenced, and the
preparation was carried out in red light (see Therman, 1938).

Exposures were made at intervals of 2 seconds.

Results.

The latency for the impulses in the optic nerve was studied
as a function of intensity by using “white light” and monochrom-
atic light at 0.470 //, 0.540 fi and 0.570 /t.

Fig. 21 shows the latency for the optic impulses at six different
light intensities of monochromatic light at 0.540 /(. The diagram
in fig. 22 shows the latency as a function of log brightness at
0.540 //. The curve shows an even monophasic course within the
range tested, at the highest intensities of which the latency
asymptotically approaches a minimum value.

The same general course in the intensity-latency curve was
obtained b)^ using “white light” and spectral light at 0,470 /i and

NEUROPHYSIOLOGY OP THE OPTIC PATHWAY.

•th4MnU<l|tt(U*v|,u tiliUUmtiUltUUUtiitisiaiiiU

Fig. 21. Oscillograph records showing the action potentials in the optic nerve of
dark adapted Hungarian frog at light intensities 1 (A), l/IO (B), I/IOO (C), 1/1,000
(D). 1/10,000 (E), and 1/100,000 (F). WaA-e length 0.310 «. Time in 1/100 sec.

0.570 fi. In each, case the course illustrated in fig. 22 was obtained,
here being no indication of any break in the curve in any single
case.

IDiscussion.

Earlier investigations of the latency of the action potentials
in the optic nerve (Adrian and Matthews, 1927 a and b, 1928,
conger eel) have already been mentioned (see page 17) as weU as
the discussion of the relation between the latency of the optic
nerve potentials and that of the retinogram (Adrian and Mat-

aVKL GUSTAF BERNHABD.

THEWS, 1927 a and b, 1928; Granit, 1933; Granit and Helme,
1939; see page 25). With tbe background of the investigations men-
tioned, the course of the curve illustrated in fig. 20 seems to show
that the prolongation of the blocking time must be primarily due
to a prolongation of the latency time of the peripheral processes.
The diagrams in figs. 20 and 21 illustrate curves obtained at
comparable intensities from a physiological point of view, where

Fig. 22. Plot of latency of impulses in the optic nerve of dark adapted Hungarian
frog as a function of log brightness. Wave length O.SiO «. Time in msec.

the latency curves asymptotically approach the minima at the
highest intensities. The curves show general agreement ae regards
their general course.

Although the results have been obtained from various subjects
and thus only offer relative comparison, they will nevertheless
give distinct support for the assumption that the prolongation of
the latency in the response recorded from the centre must be to
a great extent caused by the prolongation of the latency time of
the peripheral response.

Cruickshank (1937) pointed out the fact that the curve for

NEUROPHXSIOLOGY OF THE OPTIC PATHWAY.

the blocking time would reflect the peripheral latency ciin’-e. Ac-
cording to her experiments, the curve for the blocking time in
certain cases may show a diphasic course, and she also points out
that the two phases may be attributed to rod and cone functions
respectively.

If this is really the case, the latency curve for the optic nerve
impulses in the frog, whose eye like that of the human being is
a typical mixed rod and cone eye, should show a diphasic course.
Ko such results were obtained, however, in the experiments
described above. The relation between intensity and latency
have been investigated by using both “white” and spectral light.
In the latter case monochromatic light was used in the cone
spectrum, in the rod spectrum and at the wave lengths, at which
both elements give high response, judging by the rod and cone
curves for size of 6-wave against wave-length (Gbanit and
Muxsterhjelm, 1937; Gra>ht and Wrede, 1937). If the above
mentioned deviations in the course of the latency curve were to
be attributed to different qualities in the rod and cone elements,
they ought most probably to appear at 0.540 /t; this curve, here
as well as in the other cases, however, shows a regular mono-
phasic course.

Hartlixe’s (1938) previously mentioned investigations, as well
as Grakit’s and Svaetichin’s recently (1939) published observa-
tions on the distribution of sensitivity in the spectrum for different
elements in the retina, give interesting criteria for the functional
specificity of the optic nerve fibres. It is as yet an open question,
however, whether the different receptor functions are associated
with fibres of different diameters. If such should be the case, the
results obtained are nevertheless hardly unexpected, considering
the continuous variation in diameter of the optic nerve fibres of
the frog (Burstein and Lonnberg, 1937).

The measurements made by Holmgren (personal cominunica
tion) of the fibre size in the human optic nerve show a distribution
curve generally resembling that of the frog, with the exception o
an insignificant second hump indicating a small subdmsion o
larger fibres. Having as yet no knowledge of the relation e ween
the functional specificity of the optic fibres and their ^
impossible to say whether on account of this circumstance ere
is any reason to expect a course in the blocking time cmve w c
deviates from that of the latency of the optic imp ses m
frog.

70 CARL GUSTAF BERNTI^VRD.

These investigations do not in any way indicate a dipliasicity
in the curve for the blocking time. The results point at a regular
monophasic course of this curve.

Investigations of the Retinal Action Potential of Human
Eye with Special Reference to the Relation
between Latency of the Retinogram
and Blocking Time.

The retinal action potential of the human eye was first registered
by Dewar and M’Kendrick (1873, 1874; Dewar, 1877) and later
by Hartline (1925), ICahn and Lowenstein (1924), Sachs
(1929), and Kohlrausch (see 1931).

Cooper, Creed and Granit (1933) subjected the human retino-
gram to a closer analysis in connection with Granit’s investiga-
tions (1933) of the action potentials in the cat’s eye, and their
records give the general characteristic of the human ERG at
continuous and intermittent illumination.

The retinogram of the human dark adapted eye lacks the nega-
tive a-wave, shows a prolonged 6-wave with low amplitude (maxi-
mum 0.2 mV against 0.5 mV in the cat at similar recording), an
insignificant c-wave and lacks a definite off-effect. The latency is
long, and according to Granit et al. goes up to 60 msec, at 10 ml.
brightness. Granit et al. found in agreement with Sachs (1929)
that flickering light gave a wavy response.

Records published by Groppel, Haas and Kohlrausch (1938)
show a retinogram of a similar shape. The authors point out the
existence of certain periodical potential changes, which they
attribute to the periodical after-images.

Technique.

The subject was dark adapted previous to the experiment
and the chamber was almost dark when the electrodes were fixed.

The light stimuli were given at intervals of approximately
one minute. The intermittent lightstimuli were obtained by
a rotating flickering screen driven by a motor giving no dis-
turbances.

Altogether about 15 experiments were carried out on different
subjects and some 400 records were taken.

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY.

Results.

The record in fig. 23 A sliows the characteristics of the human
ERG., as pointed out b)” Granit et al. This retinogram which is

Fig. 23. O.'icilJograph records of retinal action potential of human eye at light
intensities 1 (A), 1/10 (B), and 1/100 (C). Standard conditions.

Time in 1/50 sec.

obtained at intensity 1, shows no sign of any negative a-wa\e.
After a latency of approximately 60 msecs, the 6-wave rises straight
from the unchanged base line. A slight secondary rise (c-wave) is
visible. At the cessation of hght, the effect falls slowly towards
the base line without giving any noticeable positive off-effect.
Eig. 24: A shows the retinal response to intermittent light at a ow

aUtL. GUSTAF BERNaVKD.

Fig. 24. Oscillograph records ol human retinal action potential. Intermittent
light at about 6 (A), 8 (B), 12 (C) and 20 (D) flashes per sec. Diameter of stimu-
lating disc about 6° 30' of visual angle. Time in 1/50 sec.

rate of alternation. After the initial 6-TVave, potential ripples,
such, as “subliminal 6-waves” follow on the general effect. Ke-
cords in fig. 24 B — illustrate the response at higher frequencies
of the flickering light. Eecord 24 C still shows distinct waves
which can no longer be identified in record 24 D. In this experi-
ment as well as in others carried out for the sake of control and
with the saiiie light intensity and area, the potential oscillations
could not be observed when the frequency exceeded 20 — 21 per
sec., while the subjective fusion frequency in these experiments
was 25 — 26 pr. sec.

The records in fig. 23 show the human retinal response at con-
tinuous illumination at three different intensities, corresponding to
the three highest of the intensities used in the experiments on the
blocking time. The different phases of the retinogram are not so
clearly marked at the low intensities; in fig. 23 C they are
still identifiable, but disappear entirely at lower intensities.
As [will be seen, the amplitude of the 6-wave diminishes with
decreasing ] hght intensity (cp. Geanit 1933, cat) at the same
time as it shows a slower rise from the base line, a circumstance
causing certain difficulties for the exact reading of the latency.
The intensity used in fig. 23 C still allows, however, of a fairly
accurate reading of the latency of the retinogram.

Table 2 shows the average values for the latency of the retino-
gram at the three different intensities (16 values at each).

NEUROI’HYSIOLOGY OF THE OPTIC PATHWAY,

Table 2.

Latency/, of the human reiinogram at different intensities.

■

Intensity

Latency of the
h-wave in 1/100 sec.

5.8 ± 0.1

1/10

7.6 ± 0.2

1/100

12.1 + 0.8

Finally, fig. 26 illustrates tlie simultaneous recording of the’
action potentials from the brain and from the retina at intensity 1,

■! »iiiii 'll _

. ■ (
' I

■; •■!

■ t . '' ! : ‘ ■

Fig. 23, Simultaneous oscillograph records of occipital alpha blooldng (upper’
cathode ray) and retinal action potential (lower cathode ray) at light intensity 1.
Standard conditions. Time in l/.'iO sec.

Discussion.

The results obtained confirm the observations of earlier authors-
conceming the general characteristics of the human retinal res^
ponse to continuous and intermittent stimulation (Granit et al.
1933; Kohlrausch 1931; Groppel et al. 1938 and others), l^at
is most striking is the long latency and the relatively insignificant
6-wave of the human retinogram. Granit et al. (1933) pomted
out that these two facts might be explained by a relatively
large negative component P III, whereas according to the same
author, the absence of a visible o-wave does not support such an
assumption.

According ’to the observations of Granit et al., these investiga-
tions also show that the fusion frequency of the potential ripp es
seems to be lower than that of the subjective. With the illuinma-
tion used, electrical fusion appears at a frequency value ma ng

CARL GUST^VF BERNHARD.

SO % of that of the subjective. This does not imply that the retinal
and sensory fusion really do appear at different frequencies
(Granit et ah, 1933). On account of the smallness of the poten-
tial waves, it is difficult to determine accurately the fusion point
of the ripples on the retinal action potential.

The lower electrical frequency may, therefore, be explained on
technical grounds, there being reason to suppose that the retinal
frequency is not lower than the sensory (Creed, Cooper and
Granit, 1933; Creed and Granit, 1933).

With regard to the absence of off-effect and the appearance of
the response to flickering light, the human retina most closely
resembles the E-type characterized by Granit (1935).

Fig. 25 shows the distribution of the total blocking time in a
“pre-retinal” and “post-retinal” period with a simultaneous re-
cording of the retinogram and the alpha blocking. As is evident
from the discussion in part I (page 25), it is difficult to determine
mth accuracy the exact time for the start of the impulses judging
by the retinogram. If the starting point of the human retinogram
not only represents the point at which the positive component
overcomes the invisible negative component but in reality the
beginning of the 6-wave, it is probable that the impulses start
after a latency of 50 — 60 msecs, at intensity 1.

The long peripheral latency proved is scarcely surprising in
consideration of the experiments on the motor reaction time at
different sensory stimuli. Even in 1898 Kichet called attention
to the fact that the relatively long reaction time at visual stimuli
in relation to e. g. that of acoustic stimuli, is to be attributed to
the slow course of the retinal processes. A comparison of the values
obtained of the reaction time at acoustic and visual stimuli taken
from works founded on more extensive investigations {e.g.
Kichet, 1898; Koga and Morant, 1923; Starling, 1930; Michon,
1939) shows a difference of a magnitude which supports the func-
tional significance of the long latency in the human ERG.

A comparison between the values obtained on the latency of
the retinogram vuth those Cruickshank (1937) gives on the
latency for the local response from the human optic cortex would
imply that the local cortical response "immediately” follows that
of the retina. The local effect suggested by Cruickshank (1937),
must, however, be regarded as somewhat uncertain. In connec-
tion with these and earlier investigations (Bernhard and Skoo-
LUND 1939) more than 50 individuals have been examined and

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY. 75

not in one single case lias any distinct, definite local effect been
proved.

If tbe impulses really start at the time when the retinogram be-
gins, it is, however, very probable that the time between the be-
ginning of the ERGr and the “hidden” response from the optic
cortex in the human brain is to be measured in milliseconds, as
indicated by the conduction rate of nerve fibres in general ( Be-
langer and Gasser, 1937), and judging by the investigations of
the response of the optical cortex in warm-blooded animals
(Bartley, 1936 b, rabbit).

It is therefore striking that the regularly appearing widespread
form of cortical reaction consisting of the abolition of the alpha
frequency appears so late. The alpha blocking does not appear
until the 6-maximum of the retinogram has been passed, indicating
that a “central period” from the “hidden” local response to the
blocking of the alpha waves goes up to more than 100 msecs.

It has already been pointed out (p. 68) that the prolongation
of the blocking time at lower intensities will be referred to a
prolongation of the latency of the peripheral processes. A com-
parison of the values for the latencies of the retinogram at those
three intensities at which it has been technically possible to ob-
tain retinograms suitable for analysis gives a further criterion for
this.

The diagram in fig. 26 illustrates schematically for these three
intensities the relation between the blocking time and that part
which represents the latency of the retinogram. The whole column
(both shaded and unshaded parts) represents at every intensity
the value of the blocking time obtained from the diagram in
fig. 20, while the shaded part gives the latency of the retinogram
from table 2. As will be seen, the prolongation of the blocking
time corresponds to the prolongation of the latency of the retino-
gram. It has previously been mentioned (p. 17) that the interval
between the beginning of the retinogram and the starting point
of the nerve impulses in experiments on animals was proved to be
constant (Adrian and Matthews, 1927 a, conger eel). Assuming
that the same applies to the human being, the prolongation of t e
blocking time at the intensities tested is due to the prolongation
of the latency, of the peripheral processes.

The relation between the blocking time and the latency ° ®

retinal response can for technical reasons only be investigate in
limited intensities on humans. A comparison between t e resu s

CARL GUSTAP BERNHARD.

.available on the human eye (table 2) and those on the frog iri fig.
26 seem to suggest that the same relation will generally hold
even for lower light intensities.

For the reasons stated above, the results can only be given with
the accuracy allowed by the time unit (1/100 sec.) used. Any
possible changes of the “postretinal” time of considerably less

Fig. 26. Diagram showing the relation between, blocking time (whole column, both
shaded and unshaded part) and latency of the retinogram (shaded part) at light
intensities 1, 1/10 and 1/100. Time in msec.

magnitude must consequently escape judgment. Nevertheless
the experiments strongly support the assumption that the blocking
time as a function of light intensity is mainly due to the peripheral
processes.

In their previously mentioned experiments on the conger eel
(see p. 8), Adrian and Matthews (1927 a and b, 1928) proved
that the latency of the optic impulses due to summative processes
in the retina is a function of intensity, duration and area of illu-
mination. It is also obvious from other observations that the block-

NEUROPHYSIOLOGY OP THE OPTIC PATHWAYL

ing time closely reflects the changes in the latency influenced by
the summative processes of the retina. Thus, it has been possible
within certain bmits to show* that the blocking time is a function
of the duration of the illumination (Cruickshahk, 1937). Again,
it is evident from experiments carried out by the author in
connection with the investigations described here that the block-
ing time is also a function of the area illuminated.

Relation between Blocking Time and Motor
Reaction Time.

Technique.

The investigations were carried out under the experimental
conditions already described (p. 61). The motor reaction and the
blocking of the occipital alpha rhythm were recorded simultane-
ously. By means of a convenient contact in the right hand, the
subject signalled when he saw the light. The contact was directly
connected with the one beam of the oscillograph. In certain series
of experiments the subject marked the moment he saw the Ught by
means of a rapid bend of the right under-arm, and then the action
potentials were registered from the biceps. Some 50 experiments
were performed on 5 subjects. At least 16 values have been
taken to form the basis af each average value in the table.

Eesults.

Fig. 27 A is a typical record with simultaneous registration of
the blocking time and the motor reaction time at intensity ^ho,
which shows how the motor response at the marking with the
hand contact takes place about 0.10 sec. after the blocking of the
alpha rhythm. The reaction time thus registered is charged with
•a time interval due to the contact and its handling. In order to
obtain further particulars as to whether the motor response rea y
does take place after the alpha blocking, the muscle action curre^s
were registered from the biceps when the subject, on seeing e
light, quite slightly and rapidly bent the under-arm. Even this
motor reaction, recorded without any mechanical loss °
took place after the blocking of the occipital alpha r ^

B). In the former case the difference of time between the motor
reaction and the blocking time was on an average 80 ±
while in the latter it was 42 rh 4: msecs, for the same m ens

CARL GUSTAP BERNILVRD.

Fig. 27. Simultaneous oscillograph records showing the relation between the
blocking of the occipital alpha rhythm and the motor reaction. A — C, hand
contact (lower cathode ray); the onset of h'ght is also marked on the lower
cathode ray. Time in 1/100 see, D, action currents from biceps (upper cathode

ray). Time in 1/50 sec.

Intensities: 1/10 (A), 1 (B), 1/100000 (C), 1 (D).

Tftblo 8.

The difference between btoching time and motor reaction time at
different intensities.

Intensity

Difference in

1/100 sec.

8.0 i O.s

1/1000

7.G + 0.5

1/10000

8.3 ± 1.-}

KEUROPHYSIOLOGY OP THE OPTIC PATHTWAY. 79

Table 3 shows the averages for the difference values between the
blocking time and the motor reaction time (when marking with
the hand contact) at three different intensities obtained on the
person, whose blocking times at the same intensities are illustrated
in table 1 (p. 65). . As is seen, the difference between the motor
reaction time and the blocking time is of the same magnitude at
the three different intensities.

Discussion.

Based on the results already described, fig. 28 illustrates sche-
matically the relation between the latency of the retinogram, the
blocking time and the motor-reaction time at intensity 1. The

Fig. 28. Diagram showing schematically the different responses from retina and
cortex in relation to motor reaction (action potentials from right biceps) to light
stimulus (int. 1). Time in msec.

6-wave of the retinogram starts about 60 msecs, after the onset
of light, and the alpha blocking does not follow until after
about another 120 msecs. The motor response under the circum-
stances mentioned does not follow regularly until after the be-
ginning of the alpha hloclcing. This latter circumstance was also
pointed out by Tbavis et al. (1937) when recording the action
currents from the right extensor digitorum communis. The
motor response taken from the biceps takes place about 40 msecs,
after the blocking of the occipital alpha rhythm.'

The experiments above (p. 76) indicate that the blocking
time at weaker light intensities is prolonged with the prolongation
of the latency of the retinogram. The motor reaction time is also
a function of light intensity (e. g. Pieron 1932), which is illus-

CARL. GUST^VP BERNHARD.

trated in table 3. Durup and Fessard (1936), having compared
the prolongation of the blocking time and the motor reaction at
different intensities, consider that the former is prolonged in com-
parison somewhat more than the latter at decreasing intensities.

This fact has not been able to be established in. these experi-
ments. At the simultaneous recording of the motor reaction
time and the blocking time, the difference between them always
proved to be constant in such a manner as is obvious from the
-typical experiments illustrated in table 3, which indicate that
the reaction time at weaker stimulus intensities is prolonged in
the same way as the blocking time.

Relation between Blocking Time and Perception
Time at Different Intensities.

The fact that the reaction time as a function of intensity is
prolonged parallel with the blocking time will show that the percep-
tion time undergoes a corresponding change; in other words, the
experiments favour a time parallelism between the blocking
time and the perception time.

In order to obtain further evidence for this, special discrim-
ination experiments were carried out, in which a strong light
stimulus was superimposed on a weaker one. In repeated experi-
ments with varying intervals between the two light stimuli, the
lowest value of time was determined between the onset of the two
light stimuli, at which the subject stated that he could not perceive
that the weaker light had preceded the stronger one. In such dis-
criminating experiments the interval between the onset of fche
two light stimuli was determined when their perceptions fused
to one. In connection with the discrimination experiments, the
blocking time was determined for the two light intensities used.

Technique.

The experiments were carried out according to the conditions
stated on p. 63. The blocking times for the two intensities to be
used in the discriminating experiments were controlled on each
occasion in about 10 records for each intensity. The exposure of the
two consecutive light stimuli was carried out by means of the
arrangements already described (p. 61) and the time between their
onset was read on the records in 1/100 second units, the time

NE^ROPHYSIOLOG'S' OP THE OPTIC PATHWAY. 3]^

marker and tke camera keing used as a recording “ckionoscope”.
The two ligM. stimuli were tested repeatedly with varying inter-
vals, and on each occasion the subject stated whether he had seen
the weaker light precede the stronger or not. The interval between
the onset of the two light stimuli was set to about the value at
which the onset of the two stimuli began to merge into one single
perception. The stronger light stimulus was in mostly maximal
(intensity 1), whereas that of the weaker stimulus was varied. All
told, about twenty experiments in total were carried out on
different persons with the use of different incensities.

Besults.

Table 4 illustrates a t 3 rpical series of discriminations from an
e^eriment in which the intensities 1/10;000 and 1 were used. The

Table 4.

(For further explanation see text.)

Interval between onset of two
stimnli (intensity 1/10000 and
intensity 1 respectively) in
1/100 sec.

The onset of the
first stimnlns
perceived: +

10 . .

Average blocking tiine at intensity 1/10000 36/100 sec.

■ , , , , , 1 17/100 sec-

Difference: 19/100 sec.

Q—t0i317

CARL GUSTAF BERKILVRD.

.weaker light stimiilus preceded the stronger at different intervals
at the different exposures given in the left column. The plus sign
in the right column indicates that the subject saw the onset of the
dim light before that of the bright. When the interval between
the onset of the two stimuli fell below 18/100 secs., that of the
dim stimulus was not perceived. The table illustrates the often
recurring fact that the subject sometimes stated that he did not

Fig. 29. Oscillograph records of the occipital alpha blocking to light at intensities
1 (A), and 1/10,000 (B) obtained in connection with the discrimination experiment
illustrated in Table 4. (Subject nr 3).

see the dim light at intervals of a magnitude which in the majority
of cases distinctly gave two perceptions. Such stray values as
these are ascribed to varying attention, occasional blinking when
the stimulus is lighted, etc. The value to be noticed in this
connection is the lowest value of the time between the onset of
the two stimuli at which the subject still perceived the onset of
both stimuli, in this case 18/100 sec.

As will be seen, this value is of the same magnitude as that of
the difference between the averages (see table 4) of the blocking
times (cp. fig. 29) obtained- in connection with the discrimination
experiment for the two intensities used.

neurophysiology of the . optic PATHW^AY'.
Table o.

Subject

Blocking times in 1/100 sec. of stimuli
at intensities:

Difference be-
j tween block-
j ing .times of

Shortest
interval
for double

nr.

1/20000

1/10000

1/1000

1/100

I tbe two stim-
j nli in 1/100
j eec.

perception
in 1/100
sec.

—

■ —

—

; —

—

. 4

—

■

—

■ —

—

Further explanation see text.

Table 5 illustrates data from 11 tjrpical discrimination experi-
ments, in connection ■vntb.tvbicb tbe blocking times for tbe differ-
ent intensities used bave been determined. The blocking times
for the two light stimuli used in each, series of experiments are
given in coluinn II. For the first, i. e. the weaker, the intensities
1/20,000, 1/10,000, 1/1000,' 1/100 were used, while the stronger, i. e.
the superimposed , stimulus, remained the same (intensity 1)
throughout. The difference of values is given in column IH,
■while, column IV gives the lowest intervals at which the subject
in each particular case clearly perceived the onset of the two
light stimuli. The table shows that in each separate case the
difference between the blocking times of the two intensities is
of the same magnitude -as .the discrimination value obtained.
They are prolonged in the same way as the intensity decreases.

Discussion.

If a strong light stimulus is superimposed on a weak one, the
onset of. the two stimuli is perceived separately, if the interval

CAKL, GUSTAF BERKHARD.

between them is sufficiently great as compared with tbeir respec-
tive perception times (tbe time from tbe onset to tbe perception
of light). If the interval between the onset is diminished, a value
is gradually obtained, in which the perceptions of the two light
stimuli, merge into one. If a fixed time relation exists between
the blocking of the occipital alpha rhythm and the perception of
light, this value should be equal to the difference between the
blocking times of the two intensities used. The results of the
experiments above described firmly support this fact within the
limits of accmacy that the time unit allows.

The values obtained in the discrimination experiments show
(see table 5, column IV) how much longer the perception times
are for the weaker light stimuli (intensities 1/20,000, 1/10,000,
1/1000, 1/100) than for the superimposed one (intensity 1). In
the same manner the values in column III show how much longer
the blocking time is for each lower intensity than for the highest.
The agreement between the values in columns III and IV shows
that the perception time with decreasing stimulus intensity is
prolonged corresponding to the increase of the blocking time.
Thus, the experiments show that a parallel of time exists be-
tween the perception of the light and the blocking of the oc-
cipital alpha rhythm caused by the light.

General Discussion and Conclusions.

The most obvious effect of the light stimulus on the human
brain potentials is the blocking of the occipital alpha rhythm,
some different interpretations of which (Berger, Jasper, Adrian)
have already been mentioned. The latency of this widespread
form of cortical response to light i. e. the blocking time, is proved
in experiments confirming earlier observations (Durup and Fes-
SARD, 1936 and 1936; Cruickshank, 1937) to be a function of
illumination intensity. This fact, as well as the circumstance that
the blocking time has been proved to be a function of light dura-
tion (Cruickshank, 1937) and area (own investigations), indi-
cates that the blocking time closely reflects the changes of the
peripheral latency which is influenced by the summative pro-
cesses of the retina (Adrian and Matthews, 1927 a and b; 1928).
The comparative investigations concerning the latency in the
optic impulses of the frog go still further to illustrate the time
relationship between the peripheral and cortical events.

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY. 85

The most direct evidence that the prolongation of the blocking
time' is chiefly- caused by -the- -prolongation of the peripheral la-
tency is obtained by recording the action potential of the hu-
man eye. Assuming that in human -beings the interval between
the retinal and nerve processes is constant (Adrian and Matt-
hews, conger eel), the prolongation of the blocking time within
the intensity range tested (see p. 73) will be ascribed to the
periphery, and there is no reason to assume that such is not the
case concerning other intensities.

Jasper (1936) based his theory for the explanation of the
blocking reaction (see p. 55) on the fact that the blocking time
is a function of stimulus intensity, a quality that Jasper con-
siders to be of cortical origin. The above investigations show,
however, that the prolongation of the blocking time with de-
creasing stimulus intensities must be primarily due to the peri-
phery. Neither have the slow potential changes maintained by
Jasper (1936) been identified. Experiments carried out with,
the use of a direct-coupled amplifier do not point to any such
regular potential changes. Both when using a D. C. amplifier-
and condenser coupled amplifier with . a great time constant.
(Jasper, 1936), an absolutely steady base line cannot be relied
upon when recording from the skin, a fact that considerably
complicates the judgment of possible potentials with slow time
characteristics.

Any definite local “evoked potential" has not been found in
these investigations. There is, however, reason to assume from
the above grounds that the local cortical response follows so
closely upon the retinal that there must be an interval of
approximately 100 msec, before the blocking of the occipital
alpha rhythm takes place. The magnitude of this “central
period" is somewhat surprising.

It may only be pointed out that the magnitude of this time
corresponds approximately to the duration of an alpha wave.
It has been proved that the alpha frequency is a function of age
(e. </. Lindsley, 1936; Smith, 1937; Bernhard and Skoglund,
1939). Thus, at the age of 1 year it makes 5 waves per sec. and
rises continuously thereafter, showing the most rapid increase
in early childhood, and then, in later youth, gradually approaches
the value of about 10 waves per sec, which is characteristic for
the adults. Bernhard and Skoglund (unpublished observa-
tions) found that the blocking time in different ages seems

86,

CAKIi GUSTAF BERNHARD.

to stand in relation to tke duration of the alpha wave. They
found that a one-year-old child, whose alpha duration is twice
as long as that of a grown-up person, also shows a blocking
time which is twice as long. The values in the ages ranging be-
tween seem to group themselves round about the curve for the
duration of the alpha waves. A correlation of the results obtained
by means of extensive investigations in the age developments
of the reaction time in the case of different senses (Koga and
Morant, 1923) indicates that the long blocking time of children’s
ages is a central phenomenon.

Whether a truly functional connection does really exist between
the central part of the blocking time and the duration of the
alpha . wave, or whether they are two variables, independently
related to e. g. the age, cannot be determined. This fact, which
has also been pointed out by Lindslew (1938) is, however,
worthy of note.

The investigations on page 80 show that there is a definite time
relation existing between the perception of light — likewise a func-
tion of intensity — and the blocking of the occipital alpha rhythm
caused by the light-. The different mental phenomena following
the presentation of a lighi} stimulus (e. g. Erismann, 1935), dis-
cussed by psychologists, cannot be entered into here, suffice to
^ay that in this connection; perception time (Empfindungszeit,
Erohlich; Verarbeitungszeit, Rubin) refers to the time from the
onset of ‘light until the time when the subject is sufficiently con-
scious of the Kght to be able to distinguish this onset from that
of a following stronger light stimulus. The prolongation of the
perception time with decreasing intensity in relation to the time
for a higher intensity corresponds to . the prolongation of the
blocking time at decreasing intensity, as compared with the same
higher intensity.

It was astronomers (e. p. Maskelyne in Greenwich, 1795)
that first began to take an interest in the value of perception
time, and in order to correlate individual astronomical observa-
tions, they endeavoured to. set up, "a personal equation” (e. g.
Bessel, 1815). Several physiologists and psychologists subse-
quently tried to go more closely into the problem (c. g. Exner,
Wundt, von Hess, Pulerich, Hazelhopf, Erohlich), and the
reaction time measurements in the case ot different sensory
stimuli have attracted great- interest (e. g. Exner, Kries, Ma-
ray, Richet, among others); results were disappointing, however.

NEUROPHXSIOLOGY OF THE OPTIC PATHAVAY. 87

^ Among later methods for the determination of perception
time, those trorked out by Hazelhoff (1923 and 1924) and
FEoHMcn ct al {e. g, Frohlich, 1922, 1925, 1929: Monj6, 1925,
1934; Vogelsang, 1925) arc worthy of note.

Irohlich’s method is founded on a phenomenon which may be
briefly described ns follows. When a vertical light slit is rapidly mov-
ed horir.ontally and appears at the vertical edge of a screen placed
between the movable light slit and the observer, the light slit is first
seen at a certain distance from the edge in the direction of the slit
movement, and not at the edge of the screen. The quotient of the
slit displacement noticed by the obscr\'er and the value of the rapid-
ity of the .slit movement is given by Frohlich ns the measurement of
the perception time. The motor reaction time is registered in the
usual manner. FRonmen ct- al. consider that this method makes it
possible to divide the total motor reaction time to light into two parts,
the ''perceptional” and the ''motorial”.

Making use of Frohlich’s method, Monje (1934, cp. Vogel-
s.\NG, 1925) found the perception time to be 33 — 40 msec, when
using n light stimulus, the total motor reaction time of which
Is about 190 msec; he also came to the conclusion that the pro-
longation of the motor reaction time at lower stimulus intens-
ities seems to be due to the prolongation of the perception time,:
MoNJfi’s value of the absolute perception time appears to be
somewhat low, considering the results already mentioned with
regard to the latency of the rctinogram at a light intensity with
corrc-sponding motor reaction time (see fig. 20 and table 2).

Frohltch’s method for the determination of the absolute per-
ception time must, moreover, be regarded with a certain amount
of scepticism on account of the criticism expressed from various
quarters.

WlRTn (1927) pointed out the improbability of li m iting the
two given parts in the motor reaction time so accurately. EttbIN
(1930 a and b)' also criticized FrOhlich’s method, and consid-
ers it incorrect .to regard the slit displacement and the perception
time as proportional magnitudes. Erissiann (1935) finally dis-
cussed the problem' thoroughly, both from a physiological and
psychological point of view.. • He calls attention to the fact that
the differences (c. g. for different light intensities) of percep-
tion time are possible to determine, but denies the possibility
of the absolute determination of perception time as a whole;
and raises strong objections to the methods of Hazelhoff and
Frohlich. .

CARL GUST^U? BERNH^UID.

Thus, an examination of recent literature gives the impression
that the contradictory opinions are the outcome of attempt’s
to solve a problem which is perhaps beyond the bounds of reason.
As a matter of faet the question arises whether it really is pos-r
sible to establish the absolute value of the perception time (see
e. g. Erismann, 1935).

In face of what has just been said, it is clear that the results
of these experiments do not aUow any conclusions to be drawn
regarding the relation between the blocking time and the abso-
lute perception time, and consequently the causal relation be-
tween the blocking reaction to light and the perception of the light
must be considered to be beyond discussion.

The results are nevertheless of interest, for they point to the
fact that an intimate time relation exists between the electrical
response of the cortex elicited by a light stimulus, and the sub-
jective perception of light.

Summary.

1. Time characteristics of the electrical responses from the
retina and the brain to light stimuli have been studied in hu-
mans. The electrophysiological data obtained have been corre-
lated with sensory phenomena following light stimuli.

2. Confirming earlier observations, the investigations show
that the latency of the blocking of the occipital alpha rhythm,
i. e. the blocking time, is a function of light intensity.

3. The results point to a regular monophasic course of the
curve representing the blocking time as a function of log bright-
ness, similar to that representing the latency of the action po-
tentials of the optic nerve as a function of log brightness ob-
tained on frog.

4. The latency of the action potential of the human eye has
been studied at different intensities, at which it has been tech-
nically possible to obtain retinograms suitable for analysis. The
experiments strongly support the assumption that the pro-
longation of blocking time is mainly due to the prolongation of
the peripheral processes in the retina.

6. Simultaneous records of the blocking reaction and the re-
tinogram show the interval between the beginning of the retino-
gram and the commencement of the alpha blocking, the magni-
tude oi which suggests a “central period” of more than 100 msec.

XEUROPHYSIOLOGV OP THE OPTIC PATHWAY. 89

The approximate correspondence between this period and the
duration of the alpha wave has been discussed.

6. Simultaneous recording of the blocking time and the motor
reaction time show that the motor response follows after the
blocking of the occipital alpha rhythm. The prolongation of the
motor reaction time with decreasing intensity corresponds to
that of the blocking time.

7. Special discrimination experiments have been carried out,
in which a strong light stimulus at varying intervals was super-
imposed upon a weaker one. The interval between the onset of
the two light stimuli, at which their preceptions fused to one,
was proved to be of the same magnitude as the difference be-
tween the blocking times of the two light intensities used. The
results point to the fact that a certain time relation exists between
the perception of light and the blocking of the occipital alpha
rhythm caused by the light.

References,

Adrian, E. D.: Proc, Eoy. Soc. 1930. 106 B: 596.

Adrian, E. D.: J. Physiol. 1931. 72: 132.

Adrian, E. D.: The Mechanism of Nervous Action. Oxford Univ.
Press. 1932 a.

Adrian, E. D.; J. Physiol. 1932 b. 76; 26 P.

Adrian, E. D.; J'. Physiol. 1937 a. 91; 66.

Adrian, E. D.: Irish J. of Med. Sci. 1937 b. 6: 237.

Adrian, E. D. and Matthews, B. H. C,; J. Physiol. 1931 a. 81; 410.
Adrian, E. D. and Matthews, B. H. C.; Bram.1931 b. 67; 355.
Adrian, E. D. and Matthews, R.: J. Physiol. 1927 a. 63; 378.
Adrian, E. D. and Matthews, R.: J. Physio). 1927 b. 61; 279.
Adrian, E. D. and Matthews, R.: J. Physiol. 1928. 65; 273.

Adrian, E. D. and Yamagiwa, K.; Brain. 1935. 58: 323.
Altenburger, H.: Bdmhe u. Foersters Handb. Neurol. II 1937, p.
717.

Barron,. D. H. and Matthews, B. H. C.: J. Physiol. 1936. 87; 26 P.
Barron, D. H. and hlATTHEws, B. H. 0.: J. Physiol. 1938. 02; 276.
Bartley, S. H.: Amer. J. Physiol. 1931. 108; 397.

Bartley, S. BL: J. cell. comp. Physiol. 1936 a. 8; 11.

Bartley, S. H.: Amer. J. Physiol. 1936 b. 117; 338.

Bartley, S. H. and Bishop, G. H.: Amer. J. Physiol. 1933. a 108;
159.

Bartley, .S. H. and Bishop, G. H.: Amer. J. Physiol. 1933. b 103;
173.

Berger, H.: Arch. Psychiat.- Nervenkr. 1929. 87; 527.

Berger, H.: Arch. Psychiat. Nervenkr. 1931. 102; 538.

Berger, H.: Arch. Psychiat. Nervenkr. 1935. 103; 111.

Berger, H.: Arch. Psychiat. Nervenkr. 1936. 104; 678.

Berger, H.: Arch. Psychiat. Nervenkr. 1938. 108; 407.

Bernhard, C. G. and Skogldnd, C. R.: Skand. Arch. Physiol. 1939.
82; 178.

Bernhard, C. G. and Skogldnd, C. R.: Unpublished observations.
Bertrand, I., Delay,. J. and Guillain, J.: L’electroencephalogramme
normal et pathologique. 1939. Masson et Cie, Paris.

Bishop, G. H. and O’Leary, J. L.: Amer. J. Physiol. 1936. 117: 292.
Bishop, G. H. and O’Leary, J. L.: J. Neurophysiol. 1938. 1; 391.
Blake, H. and Gerard, R. W.: Amer. J. Physiol. 1937. 119; 692.
Bonnet, V. and Bremer, F.: G. R. Soc, Biol. Paris. 1938. 127: 806
and 812.

NEUROPHYSIOLOGY OF THE OPTIC PATHWAY.

Bremer, F.: C. R. Soc, Biol. Paris. 1937. 124: 842.

Bremer, F.: Boll. Soc. Ital. Sper. Biol. 1938 a. 18; 271.

Bremer, F.: C. R. Soc. Biol. Paris. 1938 b. 127: 355.

Brucke, E. Th. V. and Garten, S.: Pflug. Arab. ges. Physiol. 1907.
120: 290.

Burstein, J. and Lonnberg, F.: Skand. Arch. Physiol. 1937. 76;

22 .

Cajal, S. Ramon Y.: Die Retina der Wirheltiere. Wiesbaden 1894.
Charpentier, G.; Acta ophthal., Kbh. 1936. Suppl. IX.

Claes, E.; Arch. int. Physiol. 1939. 48; 181.

Cooper, S., Creed, R. S. and Granit, R.: J, Physiol. 1933. 79; 185.
Creed, R. S. and Granit, R.: J. Physiol. 1933. 78: 419.
Crescitelli, F. and Jahn, T. L.: J. Cell. comp. Physiol. 1939. 18:
105.

Cruickshank, M. R.: J. exp. Psychol. 1937. 21: 625.

Davis, H.: Cold Spr. Harb. Symp. Quant. Biol. 1936. 4: 285.

Davis, H.: Ann. Rev. Physiol. 1939. I: 345.

Davis, H. and Davis, P. A.; Arch. Neurol. Psychiat. Chicago 1936.
36; 1214.

Davis, H., Davis, P. A., Loomis, A. L., Harvey, E. N., Hobart, 6.:
J. Ncurophysiol. 1938. 1: 24.

Davis, H., Davis, P. A., Loomis, A. L., Harvey, E. N., Hobart, G.:

J. Neurophysiol. 1939. 2: 129.

Dewar, J.: Nature. 1877. 15; 433.

Dewar, J. and M’Kendrick, J. G.: J. Anat. Physiol. 1873. 7: 275.
Dewar, J. and M’Kendrick, J. G;: Trans. Roy. Soc. Edinb. 1874.
27; 141,

Ddbner, H. H. and Gerard, R. W.: J. Neurophysiol. 1939, 2; 142.
Ddrup, G. and Fessard, A.: Annee Psychol. 1935. 86: 1.

Durup, G. and Fessard, A.: C. R. Soc. Biol. Paris 1936. 122; 756.
Dosser de Barenne, J. G. and Me Culloch, W. S.t Amer. J. Physiol.
1936. 114; 692.

Dosser de Barenne, J. G. and Me Colloch, W. S.: Z. ges. Neurol.

Psychiat. 1938. 162; 815.

Eccles, j. C.: j. Physiol. 1935 a. 85: 207.

Eccles, j. C.: j. Physiol. 1935 b. 85; 464.

Eccles, J. C. and Pritchard, J. J.: J. Physiol. 1937. 89: 43 P.
Ectors, L.: C. R. Soc. Biol. Paris 1935. 120; 1339.

Ectors, L.: Arch. Int. Physiol. 1936. 43: 267.

Einthoven, W. and Jolly, W. A.; Quart. J. exp. Physiol. 1908. 1:
373.

Erismann, Th.: Arch, ges, Psychol. 1935. 93: 253.

Erlanger, j. and Gasser, H. S.: Electrical Signs of Nervous Activity,
University of Pennsylvania Press, Philadelphia 1937.

Fischer, M. H.: Pfliig. Arch. ges. Physiol. 1932. 230: 161.

Fischer, M. H.: Pfliig. Arch, ges. Physiol. 1934. 283: 738.

Frohlich, F. W.: Z. .Sinnesphysiol. 1922. 54: 58.

Fbohlich, F. W.: Pfliig. Arch, ges. Physiol. 1925. 208: 120.
Frohlich, F. W.: Die Empfindungszeit, G. Fisher, Jena 1929.

CAKU GUSTAP BERNHAKD.

Garten, S.: Handb. ges. Augenheilk., Graefe-Saeroich. 1907. III. 1.
Gasser, H. S.: J. Physiol. 1935. 85: 15 P.

Gasser, H. S.; J. Neurophysiol. 1939. 2: 361.

Gasser, H. S. and Graham, H. T.: Amer. J. Physiol. 1933. 103: 303.
Gerard, R. W.; Cold Spr. Harb. Symp. Quart. Biol. 1936. 4: 292.
Gerard, R. W.: Scientific Monthly, 1937. 44: 48.

Gerard, R. W., Marshall, W. H. and Saul, L. J.: Arch. Neurol.
Psychiat. Chicago. 1936. 86: 675.

Goldman, G., Segal, J. and Segalis, H.: C. R. Soc. Biol. Paris. 1938.
127; 1217.

Gotch, F.: j. Physiol. 1903. 29: 388.

Gotch, F. and Horsley, V.: Philos. Trans. 1891. 182: 267.

Graham, C. H.: J. cell. comp. Physiol. 1932. 2: 295.

Graham, C. H., Kemp, E. H., Riggs, L. A.: J. gen. Psychol. 1935.
13: 275. .

Graham, C. H. and Riggs, L. A.: J. gen. Psychol. 1935. 12: 279.
Graham, H. T.: J. Pharmacol., Baltimore. 1930. 39: 268.

Graham, H. T.; Amer. J. Physiol. 1933. 104: 216.

Graham, H. T. and Gasser, H. S.: J. Pharmacol., Baltimore. 1931.
43: 163. .

Granit, R.; The Physical and Optical Societies Report of a Joint
Discussion on Vision. 1932. 263.

Granit, R.: J. Physiol. 1933. 77: 207.

Granit, R.: J. Physiol. 1935. 85: 421.

Granit, R.: Acta ophthal., Kbh. 1936. vol. 14. Suppl. VIII.

Granit, R.: Documenta Ophthalmologica. 1938. 1: 7.

Granit, R. and Helme, T.: J. Neurophysiol. 1939. 2: 556.

Granit, R. and Munsterhjelm, A.: J. Physiol. 1937. 88: 436.
Granit, R., Munsterhjelm, A. and Zewi, H.: J. Physiol. 1939. 96:
31.

Granit, R. and Riddell, H. A.; J. Physiol. 1934. 81: 1.

Granit, R. and Therman, P. 0.: J. Physiol. 1935. 88: 359.

Granit, R. and Therman, P. 0.: J. Physiol. 1937. 91: 127.

Granit, R. and Therman, P. 0.: J. Physiol. 1938. 98: 7 P.

Granit, R. and Wrede, C. M.: J. Physiol. 1937. 89: 239.

Granit, R. and Svaetichin, G.: Upsala Liikareforenings forhandlingar,
Ny foljd. 1939. 45: 161.

Grass, A. H. and Gibbs, F. A.: J. Neurophysiol. 1938. 1: 521.
Groppel, F., Hass, F., Kohlrausch, A.; Z. Sinnesphysiol. II Abt.
1938. 67: 207.

Hartline, H. K.: Amer. J. Physiol. 1925. 78: 600.

Hartline, H. K.: Amer. J. Physiol. 1928. 88: 466.

Hartlin*e, H. K.: Amer. J. Physiol. 1935. 118: 59.

Hartline, H. K.: Amer. J. Physiol. 1938. 121: 400.

Hartlinte, H. K. and Graham, G. H.: J. cell. comp. Physiol. 1932.
1: 277.

Hazelhoff, F. F.: DeWaarnemingstijd, M. de Waal, Groningen. 1923,
Hazelhoff, F. F. and Wiersma, H.: Z, Psychol. Physiol. Sinnesorg.
1924. 96: 171.

NEUROPHYSIOLOGY OP THE OPTIC PATHWAY.

Hecht, S.: Handb. gen. exp. Psycbol. 1934. Vision II, 704. Wor-
■ cester, !Mass.

Heinbecker, P.: Amer. J. Physiol. 1936. 117: 686.

Hoagland, H.: J. gen. Physiol. 1933. 17: 195.

Holmgren, F.: Untersuch. Inst. Heidelberg. 1880. Ill: 278.
Holsigren, Hj.: Personal communication.

Hughes, J.,' McCough, G. P., Stewart, W. B.: Amer. J. Physiol.
1937. 118: 411.

Hughes, J. and Gasser, H. S.: Amer. J. Physiol. 1934 a. 108: 295.
Hughes, J. and Gasser, H. S.: Amer. J. Physiol. 1934 b. 108: 307.
Ishihara, M.: Pfliig. Arch. 1906. 114: 569.

Janzen, K. and Kornmuller, A. E.: Arch. Psychiat. Nervenkr. 1939.
109: 274.

Jasper, H. H.: Cold Spr. Harb. Symp. Quart. Biol. 1936. 4: 320.
Jasper, H. H. and Andrews, H. L.: J. gen. Psychol. 1936. 14: 98.
Jasper, H. H. and Andrews, H. L.: Arch. Neurol. Psychiat., Chicago.
1938. 39: 96.

Jasper, H. H. and Cruickshank, E.. M.; J. gen. Physiol. 1937. 17:
29.

Kahn, E. H. and Lowenstein, A.: Graefes Arch. Ophthal. 1924.
114; 304. ,

Kappers, C. U. a., Huber, G. C. and Crosby, B. C.: The comparative
anatomy of the nervous system of vertebrates. Macmillan, N. Y.
1936.

Kohlrausch, a.: Arch. Anat. Physiol. Lpz. 1918. p. 195.
Kohlrausch, a.: Handb. norm. path. Physiol. 1931. Vol. .12/2.
Kopsch, F. E.: Nom. Anat., Thieme, Lpz. 1935.

Korn.muller, a. E.: J. Psychol. Neurol., Lpz. 1932. 44: 447.
Kornmuller, a. E.: Die bioelektrischen Erscheinungen der Hirn-
rindenfelder. Thieme. Lpz. 19.37.

Kubie, L. S.: Brain. 1930.- 53: 166.

Kuhne, W. and Steiner, J.:. Untersuch. Inst. Heidelberg. 1880. 3;

327.

Lehmann, J. E.: Amer. J. Physiol. 1937.. 118: 600 and 613.
Libet, B. and Gerard, E. W.: J. Neurophysiol. 1939.- 2: 153.

Lindsley, D. B.: .Science. 1936. 84: 354.

Lindsley, D. B.: J. gen. Psychol. 1938 a. 19: 285.

Lindsley, D. B.: J. exp. Psychol. 1938 b. 23: 159.

Lloyd, D. P. C.: J. Physiol. 1937. 91: 296.

Lloyd, D. P. C.: J. Physiol. 1939. 96: 118. i i looc

Loomis, A. L., Harvey, E. N., Hobart, G. A.: J. exp. Psychol. 1936.

19* 249

Loomis, A. L., Harvey, E. N., Hobart, G. A.: J. exp. Psychol. 1937.
227

Looms, A. L., Harvey, E. N., Hobart, G. A.: J. Neurophysiol. 1938.

LoRENTE^de N6, E.: J. Psychol. Neurol. Lpz. 1934. 46: 113,
Lorente de N6, E.: Amer. J. Physiol. 1935. 113: 505.
Lorente de N6, E.: J. Neurophysiol. 1938. 1; 207.

CARL GUSTAF BERNHARD.

Matthews, B. H; C.: Proc. Roy. Soc. 1937. 123: 416.

.MrcHON, P.: Le temps de reaction, Masson et Cie, Paris. 1939,
Monje, M.: Pfliig. .Arch. ges. Physiol. 1925. 208; 110.

Monje, M.; Z. Biologie. 1934. 95: 467.

Morant, 6. IVL and Koga, V.: Biometrika. 1923. 16: 346.

Obrador, S. and Odoriz, J. B.: J. Physiol. 1936. 86; 269.

PiERON, H.:- C. B. Soc. Biol., Paris. 1932. Ill: 626.

Piper, H.: Arch. Anat. Physiol., Lpz; 1911. p. 85.

Rempel, B.. and Gibbs, E. L.: Science. 1936. 84; 334.

Richet, Ch.: Dictionhaire de physiologie. 1898. Ill: 19.

Riggs, L. A.: J. cell. comp. Physiol.. 1937. 9: 491.

Rohracher, H.: Z. Psychol. Physiol. Sinnesorg. 1937. 140: 274.
Rohrachbr, H.: Pfliig. Arch. ges. Physiol. 1938. 240; 190.

Rubin, E.: Psychol. Forsch. 1930'a.''18; 101.

Rubin, E.: Psychol, Forsch. 1930 hi 18: 289.

Rubin, M, A.: J. Neurophysiol. 1938.. 1; 313.

Sachs, E.: Klin. Wschr, 1929. 8: 136.

Sjostrand, T.: J. Physiol. 1937. 90: 3 P.

Smith, J. R.: Proc. Soc. exp. Biol., N. Y. 1937. 36; 384,

Starling, E. H.: Principles of human physiol. London. 1930.
Therman, P. 0.: Acta Soc. Sci. Fenn. Nova Ser. B. 1938. Tom. II,
Nr, 1.

Travis, L. E. and Barber, V.: J. exp. Psychol. 1938. 22; 269.
Travis, L, E. and Hall, M. E.: J. exp. Psychol. 1938. 22: 472.
Travis, L. E., Knott, J. R. and Grepfith, P. E.: J. gen. Psychol,
1937. 16r 391.

Umrath, K.: Pfliig. Arch. ges. Physiol. 1933. 233: 357.

Umrath, C, and Umrath, K.: Pfliig. Arch. ges. Physiol. 1934. 284:
562.

Vogelsang, K.: Pfliig. Arch. ges. Physiol. 1925. 208; 93.

Wang, G. H.: Chin. J, Physiol. 1934. 8: 121.

Wang, G. H. and Lu, T. W.: Chin. J. Physiol. 1936. 10; 149.
Westerlund, a.: Skand. Arch. Physiol. 1912. 27: 260.

Whetteridge, D.: J. Physiol. 1937. 89: 99.

Wirth, W.: Handb. norm, path. Physiol. 1927. 10; 598.

Wrede, C. M.: Skand. Arch. Physiol. 1937. 77; 93.

Wright, W. D. and Gbantt, R.; Brit. J. Ophtalm. 1938. Monograph
Supplement IX.

ACTA PHYSIOLOGICA SCANDINAVICA

VOL. 1. SUPPLEMENTUM II.

From the Ph)'siological Department of the Universit}' of Lund,

Sweden.

ON THE

CITRIC ACID METABOLISM
IN MAMMALS

Johan Martensson

This paper appears in

ACTA PHYSIOLOGICA SCANDINAVICA

as Supplement II to Vol. 1.

Preface.

T his invesligalion has been conducted at the Physiological
Institute of Lund, and was commenced in 1937. The Head
of the Institute at that time, Professor Torsten Thunberg,
introduced me to this field of inquiry and provided me with
exceptionally good working facilities. Since then he has encour-
aged my work with the greatest interest and kindness, for
which I tender him my respectful and hearty thanks.

To the present Head of the Institute, Professor Georg
Kahlson, I proffer my sincere thanks for the favourable work-
ing facilities I was still allowed to enjoy and for his interested
help.

To Professor Gunnar Ahegren I also owe my cordial thanks
for kind help and for the instructive years of work I had at the
Pharmacological Institute.

I also thank Miss Margareta Brorsson for her good technical
help in the laboratory, Mrs. Karin Weber-Berg for help with
the preparation of the manuscript, and Mr. Bert Hood for the
translation of my manuscript.

The investigation has been conducted with the aid of generous
financial assistance from the Nordisk InsuUnfond, for which I
am deeply grateful.

To my wife, who carried out all the exacting work connected
with the citric-acid estimations and assisted me with the animal
experiments, I owe my warmest gratitude for her great, self-
sacrificing work and her good support throughout the entire
investigation.

Lund, in May, 1940.

Johan Mdrtensson..

Contents.

Page

Preface 3

Inlroduciion 7

Chapter I. Contribution to Thunberg's method of estimating citric acid 9

Tlie citrate curve 9

Tlie enzyme slabilitj* and spontaneous donors in cucumber-seed

extract 13

Tlic citrate curve at different hydrogen-ion concentrations 20

Influence of light on rate of decoloration 24

The form of the X-curve in relation to the standard curve 26

The systematic deviation of the serum-citrate curve 28

Calculation of the Ci-values 31

The accuracy of the enzymic method of determining Ci 32

Comparative investigations with the pentabromacctonc method .... 33
Summary 35

Chapter II. The citric acid metabolism in the normal mammalian body 37

Review of previous investigations 37

The citric-acid content of the serum after intravenous administra-
tion of citrate 39

The citric-acid content of the red blood corpuscles 41

The citric-acid metabolism in perfused isolated liver 43

The citric-acid absorption from the intestine 46

The citric-acid content of scrum after functional elimination of the

liver and portal area

The citric-acid metabolism in an eviscerated preparation and in

perfused muscle

The citric-acid excretion in the urine 33

The citric-acid metabolism in perfused kidney 33

Tlie citric-acid content of serum after nephrectomy 39

The citric-acid concentration in various tissues normally and after

administration of citric acid

Discussion of the experimental results

Summary

Chapter III. What is the cause of the hgpercitricaemia in lesions of the

hepatic parenchyma?

The Ci-metabolism after functionally cutting off the liver and

• 77

preventing absorption from the portal region

JOHAN MARTENSSON

Pago

The Cl-melabolism after funclioiially cutting off the liver but retain-
ing the portal circulation intact 78

The Ci-metaholism in experimental lesion of the liver 80

Is the Ci-metaholism in vivo connected with the amino-acid

metabolism? 82

Summary 87

Chapter IV. Some experiments on the Ci-metabolism attending changes

in the acid-base equilibrium 88

Summary 92

References 94

Introduction.

T he important part played by citric acid in the intermediary
metabolism of the body has been brought out with ever
increasing sharpness in the investigations of recent years.
These investigations have been chiefly based upon in-vitro
experiments on minced tissues or enzyme preparations from
these tissues. The present work is mainly an experimental
inquiry into the normal in-vivo metabolism of citric acid in
the mammalian body. In this work full advantage has been
taken of the increased facilities offered by Thunberg s enzy-
mochemical method for the micro-estimation of citric acid.
The inquiry has also been extended to some experimentally
induced pathological conditions, with a view to elucidating the
mechanism behind those changes in the citric-acid conten o
the serum which have been found in various morbid con i
tions in man, especially in injuries to the liver parenc
and which are utilized in the differential diagnosis of hepatitis

and obstructive jaundice (SjOstrOm, 1937).

The work was begun in October, 1937. As frequent discus-
sions of the citric-acid question were appearing in t e pijsio
logico-chemical literature, I considered it necessary ®

a preliminary account of the results I had obtaine (

SON, 1938), this being followed the next year
tion of a special problem attracting interest ^ ®

TENSSON, 1939). These results have been further
by later experiments and the work has een

embrace new fields of inquiry. The ^ -nresented

given here, on the whole, in the order t e pro . ^

themselves in the course of the work or as P^^^toents
publications from other centres. Experieimes an toaeth-

bearing on methodological questions have been br g
er in a special chapter. For reasons of ®P^® , having

from giving a review ^ 3 ^^ GrO^ale, 1937).

been done a few years ago (bJOSTROM, i

JOHAN mArTENSSON

Recent as well as earlier literature relating to problems dealt
with here are reviewed in connexion with the respective experi-
ments. A more general review of the occurrence, conversion
and significance of citric-acid in the body has also been re-
cently submitted by the author (MArtensson, 1940).

CHAPTER I.

Contribution to Tluinberg's Method of
Estimating Citric Acid.

T hunberg published his enzymic method for eslimaling
cifric-acid (abbreviated “Ci”, which also refers to citrate)
in 1929, and gave a detailed description of it in 1933. Since
then all the investigators who have made considerable use of
the method have contributed to making it more exact or more
practical for various purposes. (Ostberg, 1934; Lenner, 1934:
SCHERSTfeN, 1936; SjOstrOm, 1937; and GrOnvall, 1937.) In
this work I have as a rule used the method such as it has been
elaborated by this accumulated experience. I submit here my
own experience of the method and my tests bearing on methodo-
logical questions only in so far as they offer something new
of practical interest or as they are likely to assist in clearing
up some previously unclear aspect. Further, reference has
been made to some recent Ci studies that have given results
important for the enzymic method of estimating Ci.

The Citrate Curve.

In the following account it is assumed that the details of the method
arc known. In the discussion on the so-called Ci-curve the termino o^
usual in the Ci literature is used: the phosphate extract of cucumber-see ,
the “enzyme solution”, docs not contain only Ci dehydrogenases ut a so
other dehydrogenases and their substrates, the spontaneous donors ,
hence a distinction can be drawn between ‘ Ci activity and spontaneo
activity”. The citrate curve consists of the ”level line” and an ascending
leg”, which meet at the “contact point". The curve for the unkno^vm
solution, the “X curve”, is compared with the “standard curve .

SjGstrGm (1937) has submitted the citrate curve to an ex-
perimental analysis, in which he shows the great eanng
spontaneous donors have on the appearance o re curv

JOHAN MaRTENSSON

is however not clear from his account that the form of the Ci
curve is entirely (except in the spontaneous-decoloration tube)
determined by the co-action of the Ci activity and the spon-
taneous activity. And his statement that, medially to the con-
tact point, the spontaneous donors “steal” most of the methy-
lene blue (MB), leaving only small fractions at the disposal
of the Ci, gives an inaccurate idea of what happens. On the
contrary, the fact is that proportionally to MB there is a
shortage of Ci there, so that the latter is exhausted rather early
and the Ci dehydrogenase clearly cannot decolorize more than
a part of the MB. The rest is dealt with by the spontaneous
activity, and the decoloration-time becomes longer the less Ci
the tubes contain. It may be said that the ascending leg of the
curve gives an inverse, magnified picture of the amount of Ci
in the tubes medial to the contact point. The height of the
level line is determined by the Ci activity and spontaneous ac-
tivity jointly. The position of the contact point is laterally
determined by the ratio of the Ci activity to the spontaneous
activity: when spontaneous activity is relatively feeble the Ci
dehydrogenase is responsible for almost the whole of the MB-
decoloration, a larger amount of Ci is used up for this, ahd the
contact point lies further laterally. The slope of the ascending
leg, also, is determined by the ratio the Ci-activity hears to
the spontaneous activity. As in Ci determinations the same
enzyme is used both for the solution with unknown Ci-content
and for the standard solution, the spontaneous activity is equal
in both cases, and it is consequently the Ci-content of the un-
known solution which entirely determines the form of the curve
relatively to that of the standard curve (provided the solution
is a pure Ci one or so diluted that other disturbing substances
do not affect the reaction.

According to the above account the two straight legs of the Ci curve
should meet at a contact point. But actually the contact point is almost
always more or less non-existent, since one of the legs merges arcuately
into the other. This is because the Ci-aclivity successively diminishes
(under the law of mass action) after the Ci has come below the optimum
concentration for the enzyme. The level-line does not start at that point
in the tube series where the quantity of Ci is just in e.vcess of the MB, but
where the Ci-excess is so large that the amount of Ci required for the
reduction of the MB does not bring the Ci-conccntration below the op-
timum. In the usual arrangement of the experiment i> '/ (micrograms) of Ci
correspond to the added quantity of MB, 10 7 , (in 2 cc.) but the optimum con-

ON THE CITRIC ACID METABOLISM IN MAMMALS 11

^eiilrnlion is probably n lifUc higher. According to Dann (1931), the
Michaclis constant (the conccnlralion for the semi-maximum conversion
according to Davies and Quastel, 1932) for Ci-dchydrogcnase from
cucumber-seed is S.io" Mol. at 35° C, but his computation is doubtless
somewhat uncertain. According to Adler, v. Euler, GUnther and Pless
(1939), this constant for isocitricodchydrogennsc from heart muscle is less
than 1.25 . 10"“ Mol. and it is probably of the same order of magnitude for
the cucumber-seed extract. Another reason for the ascending leg not
being straight but curving upwards is that the spontaneous activity
gradually declines in the course of the c,xpcrimcnt, this, too, being due to
the concentration of the substrate gradually falling below optimum.

Here the fad has been ignored Uial Ihe so-called citricode-
hydrogenasc is actually more complex than was known at the
outset. Wagner-Jaubegg and Rauen (1935) showed that iso-
citric acid is also dehydrogenated by cucumber-seed extract,
even more rapidly than Ci; hence it may be an intermediate
product in the breakdown of Ci. Later on, Martius (1937,
1938) submitted good evidence that the breakdown of Ci is
initiated as follows: Ci is first deprived of water by a hydratase
(similar to fumarase but not identical Avith this), which has
been called aconitase, and is converted into the unsaturated
cis-aconilic acid. This lakps up water again and forms iso-
ciiric acid, which is afterwards dehj'drogenated by the iso-
citricodehydrogenase. Explained on these lines the reaction
accords better with the general view of dehydrogenation reac-
tions, whereas a direct attack on Ci would have been difficult
to explain. Thunberg (1929) conceived of an initial re-ar-
rangement in such manner that water was taken up with
formation of an ortho-acid, which could then be attacked by
the dehydrogenase. Tims the enzyme system consists of
aconitase and iso-citricodehydrogenasc along with co-dehydrose
II (Adler, v. Euler, GOntheb and Pless, 1939) and flavopro-
tein (Wagner-Jauregg and Rauen, 1935).

AnderSSON (1933) found an elevated Ci-dehydrogenation to result from
addition of Co-zymase, something that ScufeRSTEN (1936) was unable to
confirm. The different results obtained by these investigators may be due
to their Co-zymase preparations having had Co-dehydrase II admixtures
of different strengths, co-zymase being according to ADLER an ot ers
(1939) completely inactive as against iso-citricodehydrogenase from amma
tissues as Avell as from higher jJants and yeasts.

The objection might therefore be raised to the enzymic
method of estimating Ci that tlie rate of decoloration is no

JOHAN MArTENSSON

determined by the iso-citricodehydrogenase but by the amount
of Co-dehydrase II and flavoprotein. The cucumber-seed
extract admittedly contains a certain amount of these co-en-
zymes, although I have not yet had an opportunity of testing
whether they occur in optimum quant%. Otherwise the results
might be influenced if these co-enzymes were supplied with the
fluid to be tested in such quantity that the decoloration-time
was thereby accelerated. According to v. Euler and Sghlenk
(1939), co-zymase is present in the blood almost exclusively in
the corpuscles. Co-dehydrase II is present in less quantity than
co-zymase. And, moreover, as it is rather soon destroyed
(through phosphatase activity) if the blood is allowed to stand,
the risk of getting erroneous values through adding co-dehyd-
rase II with the blood serum ought to be very small. With
regard to the flavoproteins, of which the diaphorase ^ is pri-
marily concerned here as the co-enzyme dehydrogenase oc-
curring in the animal body (Dewan and Green, 1937, Adler,
V. Euler and HellstrOm, 1937, Straub, 1939, quoted from
Dixon, 1939), this seems to be present in all animal tissue,
but I have not been able to find any reports as to its presence
in blood serum. An assurance that the presence of co-enzyme
in the fluid to be tested cannot cause an error of any conse-
quence in the Ci-estimation is offered by the fact that the level
of the X-cmve is never found to lie with certainty below that
of the standard curve: this ought to occur if the tested fluid
contains co-enzymes and the enzyme solution has not optimum
concentration from them.

Martius’ decomposition scheme lias been confirmed by Breusch (1937)
and Adler et al. (1939), who find that in experiments with Ci tlic decolor-
ation-time becomes shorter if Ci is incubated with, the enzyme for some
time before the other components arc added; during this time the con-
version to isocitric acid lakes place. I have repeated the experiments wilb
cucumber-seed extract without finding any shortening of the decoloration-
time if Ci had stood with the enzyme under anaerobic conditions before
the addition of MB. This applied to both Ci-solution and serum. A differ-
ence of some importance might be conceivable if the Ci and isocitric
acid in the scrum are in equilibrium from the start, whereas in the case
of a pure Ci-solution this equilibrium is rot established until after addition
of the enzyme.

Andersson (1938) has questioned whether the Ci added in the cx-

^ Diaphorase II according to Abraham and Adler (1940), who consider
that they have demonstrated the existence of two diaphorascs.

ON THE CITRIC ACID METABOLISM IN MAMMALS

pcrimcnts really nets ns a substrnlc. He thinks that it may just ns well
be interpreted ns nn aclivnlor after the fashion of the Ci-dicarboxylic
acids, and is inclined to sec a proof of this in the small Ci-amounls
required for the dccolornlion. It has probably escaped his notice that only
verj- small amounts of MB nrc concerned ns well; actually about equi-
inolccnlnr quantities of Ci nnd MB arc involved, as Tm/NnERG already
pointed out in 1929.

The Enzyme Stability and Spontaneous Donors in
Cucumber-seed Extract.

In preparing the cucunilier-seed extract I have, on the whole,
followed ScHER-STtN’s (1936) directions, i. e. adopted long-time
extraction in the refrigerator. However, I have devoted some
experiments to the quc.stion of what takes place during the so-
called self-purification of the enzyme extract and to the ques-
tion of the spontaneous donors and the stability of the enzjTne.

Schersten’s experiments show that Ci-dehydrogenase is
stable for a long time if the cnz 5 'mc is ice-cooled, and tliat tlie
.spontaneous-donor systems are far less stable. Respecting the
conditions at higher temperatures his account does not give
definite information. .Sciiersten presumes that the lowered
spontaneous activity associated with long extraction-time may
be due to an exhaustion of the .spontaneous-donor systems.

If Uie same enzyme preparation is tested after extraction-
limes of different length it is found that the spontaneous activ-
ity declines successively, while 'the Ci-activity keeps fairly
constant (Fig. J).

This decline in the spontaneous activity is due to exhaustion
of the spontaneous-donor substances, not to inactivation of the
dehydrogenases attacking these substances.

Experiment 3t. 1.40. Exlmction with .M/15 phosphale buffer of pH 7.38
in room Icmpcralurc, 30 min., wlicrcupon the e.xlract was centrifuged and
te.sicd for activity: .spontaneous decoloration 16 min.; with 10 y Ci 13 min.
Pari of the solution, portion I, was boiled, filtrated, and diluted to its
original volume. Tlic rest of the enzyme solution was allowed to stand
90 min. at room temperature, whereupon another part, portion II, '"as
boiled, and the remaining enzyme was used for the experiment, the tubes
were charged with Ci, portions I nnd If respectively. Decoloration-times:

See Fig. 2. . • j.

Portion I contained the stable donor substances in the enz}me wit a
spontaneous dccolornlion-timc of 15 min., portion II the donors wit i a
spontaneous decoloration of 130 min. That the dehj’drogenase actm y

JOHAN MARTENSSON

was none the less in the same good condition could be seen from the fact
that an addition of 1 cc. of portion I gave the 15 min. decoloration-time
again.

SCHERSTEN (1936) found that the activity of the enzyme so-
lution becomes less if the extraction-time is reduced to 15 min.
or less, while the spontaneous-donor content becomes propor-
tionately larger. The cause of this was assumed to he that
the spontaneous donors enter in solution more readily than the
enzyme, or possibly that during the longer extraction-time the
spontaneous donors are used up more. That this latter as-
sumption is correct was shown by the preceding experiment.

F/ff. 1. Cucumber-seed extract with M/20 KsHPO* is divided into tbrce
portions extracted in room temperature during 1, 2 and 3 hours respectively.

But the former hypothesis is undoubtedly correct as well, it
being possible to obtain the same type of enzyme solution
even with long-time extraction if the cucumber-seeds are
pounded and stirred less thoroughly: then the enzymes — Ci-
dehydrogenase as well as the others — dissolve with difficulty,
whereas the donor substances enter readily in solution, in
which they are only slightly oxidized on account of the low
enzyme activity.

Experiment 31. 10. 38; Extraction with M/20 KsHPOt one hour in
refrigerator and 45 min. in room temperature.

Decoloration-limes:

Ci in tubes 0 2 3 4 5 6 8 y

High degree of stirring 44 25 20 14 12 12 11 min.

Low „ „ „ 23 21 20 20 20 20 20 „

ON THE CITRIC ACID METABOLISM IN MAMMALS

Experiment ii. 10. 38; Extraction with M/20 K 2 HPO 1 overnight in

refrigerator.

Ci in tubes 0 2 3 4 5 6 8'/

High degree of stirring — — 61 49 22 14 11 min.

Low „ „ „ 16 12 10 8 8 8 8 „

The experiments reveal the technically important fact that
the stirring of the cucumber-seed extract must always be car-
ried equally far if standard curves of a constant form are to
be obtained.

Fig. 2 . MB decoloration by donor substances in boiled cucumber-seed

extract (see text).

From some tests of the enzyme stability at 35° (the tempera-
ture at which Ci estimations are usually made), Scherstei
infers that the Ci-dehydrogenase is more rapidly
the spontaneous donors are reduced too much, 1 . e. ®
taneous-donor system is thought to have a protec ive in
on the Ci-dehydrogenase. In one case the enzjme v
activated so much in 30 minutes that the deco ora ion
6 r Ci rose from 16 lo 70 minutes. This however, mig. smplj
be explained thus, that this quantity of i was ®
spontaneous activity was high) just m excess ^
quantity of MB but was afterwards ^^d

vity fell) insufficient; in such cases nicumber-

great changes. The Ci-dehydrogenase actwity

seed extract remains a long time unc “”3 rninimum
even if the spontaneous activity is reduced to a minimum.

JOHAN MARTENSSON

Experiment 9.2.39. Extraction from cucumber-seed for 13 hours. in
refrigerator with M/20 KjHPOi. After being centrifuged the enzyme
solution was ice-cooled 30 minutes and a first series was then set up,
whereupon the enzyme was placed in a 35° water-bath. At intervals of
30 minutes other series were then set up. The decolorizing times are
given in Fig. 3.

The experiment also gives an idea of the successive decrease
in intensity the spontaneous activity always undergoes during

Min.

Fig, 3. Change in citrate-curve resulting from storage of the enzyme

at 35° C.

the decolorization. When the spontaneous reduction time, as
in the first series, is about 90 minutes, the dehydrogenation
intensity at the end of the decolorization should be only half
that at- the start (since the spontaneous decolorization takes
double the time after the enzyme has had a temperature of
35° for 90 min.). Thus the intensity of the spontaneous ac-
tivity at the beginniny of the experiment cannot be calculated
simply as the inverted value of the decoloration-time; it is con-
siderably greater. The increased decoloration-time for the un-

ON THE CITRIC ACID METABOLISM IN MAMMALS

doubtedly optimum Ci-quantities in the above experiment
(from 12 to 18 min.) can therefore no doubt be attributed to
diminution of the spontaneous dehydrogenation.

Doubling the spontaneous activity by adding twice the amount of en-
zyme without other addition in no wise means merely a halving of the
decoloration-time, but a considerably greater shortening of this time. In
experiment 2. 2. 40 the spontaneous decoloration took 87 min. with 0.5 cc.
of enzyme, but 20 min. with 1 cc. of enzyme, i. e. less than one-fourth.
In experiment 31. 1. 40. it was 130 min. with 0.5 cc. of enzyme, 41 min.
with an addition of 0.5 cc. of boiled enzyme extract (i. e. without enitymic
activit}', onl}’ containing the spontaneous donors), thus less than one-
third. If one adds a substance that accelerates decoloration, one is there-
fore apt to ascribe to this substance too great a participation in the in-
creased dehydrogenation activity. Thunberg (1936) has given a thorough
analysis of the factors to be observed in calculations of the kind in question.
He has none the less used a “minute decolorization value’’ derived from
the spontaneous-decoloration time, pointing out that the calculation is not
quite correct but may yet be of use in practice.

An oxidative process is thus constantly going on in the enzyme solution,
but it is of very low intensity if the solution is kept effectively cooled
down. (That the enzyme solution, as earlier investigators have pointed
out (LENN’fen, ScHERSTfiN), is most labile during the first 30 min. after
centrifugation, is probably due most to the fact that the solution has not
had time to become properly cooled down.) None the less, if for some
reason a tube fails at the beginning of a large series, a certain amount of
error is to be expected if a /resh tube is loaded and evacuated at the end
of the series and inserted' in the place of the old one.

Should the enzyme solution be found to be “too strong” on
account of excessive spontaneous activity, a suitable citrate
curve will not be obtained by diluting the enzyme, since this
also reduces the Ci-activity, but by keeping the solution at room
temperature for some time. Still, long-time extraction in a re-
frigerator probably offers better prospects of getting equivalent
enzyme preparations on all occasions than does short-time ex-;
traction at room temperature.

If the spontaneous activity is very high, the entire Ci-cuj;ve
becomes a horizontal line, which means, that the decoloration-*
time cannot be shortened by addition of Ci. This can scwcely
be mterpreted otherwise than that the enzyme solution itself
contains Ci in optimum quantity from the start,, yrhich, must
be dehydrogenated before the typical citrate-curve can be
obtained by adding Ci. This is by no means unreasonable, ifoi;
it merely presupposes that the enzyme solution contains 20 7

JOHAN mArTENSSON

of Ci per cc., that is to say, that the Ci-content of the cucumber-
seeds is about 0.02 %. The great activity of the Gi-dehydro-
genase quickly reduces the content to suboptimum concentra-
tion, 10 7 of Ci per cc being probably converted in 10 — 16 mi-
nutes at 35°. (The previously described experiment with ad-
dition of boiled enzyme agrees well with this: boiled enz5mie
having a high content of spontaneous donors gives a fairly
tj’^pical citrate-curve. Fig. 2). This was later confirmed in di-
rect estimation by the pentabromacetone method: one phos-
phate extract prepared in the usual way was analysed after it
had been centrifuged and found to contain 17 y per cc., in
another the content was 20.4 y- per cc., -which had fallen to 10.<;
7 per cc. after the enzyme had been kept at room temperature
for 90 minutes. Direct extraction from the cucumber-seeds
with trichloracetic acid yielded a quantity equivalent to 350 7
of Ci per gm. of cucumber-seed (in another case about 400 7
per gm.) This may mean that Ci plays some part in the
metabolism of the seeds, and it affords a reasonable explana-
tion of the abundant Ci-dehydrogenase content of the cucumber-
seeds.

Tliis condition also explains -why the spontaneous activity, if it is very
powerful, cannot be paralysed by ZnCl- (ScherstCn, 1936): if it depends
on Ci, it can obviously only be paralysed by ZnCb when the latter is in
such concentration as ■\\nll also inhibit the Ci-activity.

With reference to the other substances that are probably
contained as spontaneous donors in the cucumber-seed extract,
I have conducted some experiments on the dehydrogenation of
malic and hexose-diphosphoric acids (Preparations used: 1-
malic acid, British Drug Houses, and Hexosediphosphorsaures
Calcium rein “Bayer”). The decoloration-curves for these sub-
stances are in good agreement with those given by Thunberg
(1929 a and b). For hexose-diphosphoric acid the “level” is
lower than for Ci, for malic acid generally somewhat higher
than for Ci. But for these substances- there is, compared with
Ci, quite a different proportion between the plainly active
concentration and the optimal one: a series of tubes loaded
with a solution of these substances in the same quantities as
are used for Ci gives a considerably more “horizontal” decolo-
ration-curve.

ON THE CITEIC ACID METABOLISM IN MAMMALS

Experiment SI. V2. 39:

Ci 2 3 4 5 V

Millie acid 0.2 O.a 0.4 0..'> mgni.

Dccoloration-Hinc fiO 55 42 47 28 43 20 40 min.

Experiment 2. 2. 40:

Ci 2 h 10

Hcxost'-diplio.sph 4 10 100 ■/

Decoloration-lime 45 35 15 17.5 12 7.5 min.

These suhstnnees alone can therefore scarcely simnlntc Ci, though
their admixture to a Ci-solution may cause a certain amount of error in
the estimation by making the ascending leg of the curve less sleep. So
far ns malic acid is concerned, however, so great a concentration is
required that it can hardly occur other than in tolerance tests with malic
acid. On the other hand, hexosc-diphosphoric acid, even in the same
concentration as Ci, can play a certain part, but when a large quantity of
solution is added the decoloration-lime will be shorter than that represented
by the level of the standard cnn'c. This ought to arouse suspicion, control
tests with addition of monoiodoacctic acid can then decide the matter.
However, hexosc-diphosphoric acid is probably not present in the scrum
in demonstrable quantity, no more than it is in freshly taken mammalian
muscle (DnUTiCKK and Hou-man.v, 1939).

Min,

Fig. 4, Decolorizing inlensily in cuciiniber-sccd extract, stored at 35°,
after addition of hexosc-diphosphoric acid O.i mgm. (I), citric acid
0.01 mgm. (II), and malic acid 9 mgm. (III).

.Some tests -svith enzyme that had been kept at 35° showed
lliat the intensity of the malicodchydrogenation declined in a

JOHAN mArTENSSON

much higher degree than that of hexose-diphosphoric and of
especially the Ci dehydrogenation.

Experiment 3. 2. 40 and 5. 2. 40. Extraction from cucumber-seed with
M/15 phosphate of pH 7.38 overnight. After the enzyme solution following
centrifugation had been ice-cooled for 30 min., a first series was set up,
whereupon the enzyme was placed in a water-bath at 35°. Other series
were set up at intervals of 45 min. Fig. 4 shows the decoloration-times
(means of two experiments). A last series with enzyme that had been
kept ice-cooled the whole time had the same times as the first.

The result need not imply that the malicodehydrogenase is
more thermolahile than the other dehydrogenases, the inacti-
vation being possibly also due to a lowered co-zymase content.
According to Andersson (1938), for optimum activity the mali-
codehydrogenase requires a considerably larger co-zymase con-
tent than the hexose-diphosphoric dehydrogenase, and hence
any decrease of this content in the extract would primarily af-
fect the malicodehydrogenation.

The Citrate Curve at different Hydrogen-ion
concentrations.

Andersson (1938) has studied the activity of Ci-dehydro-
genase in cucumber-seed extract in phosphate buffers of dif-
ferent pH between 7 and 8. He foimd that the activity is maxi-
mum above 8 and declines very rapidly with falling pH, being
entirely inhibited at pH 7,2; hence at lower pH added Ci should
have no effect on the decoloration-time. Andersson, on the
basis of this finding, raises grave objections to the enzymic
method for estimating Ci. In this he appears to rely on only
one experiment with old enzyme extract and does not seem
to have studied the literature on the subject.

Dann (1931) investigated the activity of the Ci-delwdrogenase in- cucum-
ber-seed extract (with optimum Ci-addition) and found the optimum at
pH 7 — 10; below pH 6.5 the activity fell quickly, reaching minimum at
pH 3.5. Adams (1931) found similar conditions for an aqueous extract
of cucumber-seed: the difference in activity was small between pH 6.2 and
8.3, in more acid solution the activity was less and at pH 4.6 appeared to
he altogether inhibited. Ostberg (1931) published experimental series with
the same decoloration-times within the range pH 7.15 and 7.55. He got
some degree of difference when the pH was, on one hand, below 7
(9.5 min.), on the other hand, between 8.15 and 7.40 (6 min). Some workers
have also studied the breakdown of Ci at different pH in metabolic ex-

ON THE CITRIC ACID METABOLISM IN MAMMALS

perimenls: Laxgecker (1934) found in minced liver tissue no difference
in Ci decomposition between pH 6.9 and 8.2. In similar experiments
ISHiHAUA (1938) found only a slight difference between pH 6.2 and 8.2.
Krebs and Eggleston (1938) had good Ci metabolism in pigeon muscle
at pH 6.8 in phosphate buffer.

For isoCi-dehydrogenase from heart muscle Adler, v. Euler, Gather
and Pless (1939) got the same pH-curve as for other dehydrogenases:
high activity between pH 7 — 7.5 and rapid fall below pH 6.5. They used,
however, veronal-glycine buffer, having found that phosphate ions arrest
the activity and that this arrest varies considerably with the pH: with
phosphate buffer the curve has its maximum between pH 6 and 6.5, where
the arrest is relatively small, and its minimum between pH 7 and 8, where
the arrest is great.

I have repeated the tests on tlie sensitivity of cucumber-seed
extract to pH fluctuations within the ranges concerned here.
Like earlier, workers, I found that the activity of the Ci-de-
hydrogenasc varies very slightly, if at all, within the range pH
6.6 — 7.6, whereas the spontaneous activity rapidly declines with
falling pH and ceases almost entirely below pH 7.

The following experiments may be adduced:

1. (13. 1. 40) Two cucumber-seed preparations, 13-hour extraction in

refrigerator with M/15 Sorensen’s phosphate buffer of different pH. (pH
measured after the tests, after MB had been recoloured, with a quin-
hydrone electrode and audion valve potentiometer.)

Ci in tubes 0 2 5 10 '/

pH 7.53 58 28 15 12 min.

pH 7.10 > 240 148 20 18 „

2. (22. 1. 40) Extraction from cucumber-?eed Mith M/20 KiHPOi at
room T.; so spontaneous activity was very low. In tubes M/15 phosphate
buffer of different pH. In every tube undoubtedly optimum Ci-quantitj,
10 7 ; so the decoloration was practically speaking a measure of the Ci-
nctivihj.

pH 6.04 7.23 7.54

Decol. time . ; 20 18 17 min.

3. (22. 1. 40) Extraction from cucumber-seed with M/20 K 2 HPOj in
refrigerator short time, great spont. activity. In each tube suboptimum
amount of Ci, 3 7; so decoloration time was a relative measure of the
spontaneous activity.

pH 7.00 7.12 7.28 7.34 7.71

Decol. time 89 55 48 43 36 mm.

JOHAN mArTENSSON

(Tests with unbuffered aqueous extract from the seed showed no com-
plete spontaneous decoloration because the reaction was displaced to
below pH 7 in the course of the experiment.)

4. (1. 2. 40) Extraction from cucumber-seed with M/30 phosphate
pH 7.88. In the tubes M/15 phosphate to different pH. In each tube
l-malic acid brought to M/75. Little spont. activitj*. (> 110 min.)

pH 6.85 7.07 7.18 7.32 7.42 7.55

Decol. time 31 27 25 22 21 18 min.

TJie pH curve for the malic acid does not argue against part of the
spontaneous activity depending on a nialicodehydrogenation.

5. (9. 2. 40) Overnight extraction from cucumber-seed with M/15
phosphate, pH 7.38. In tubes 80 y hexose-diphosphoric acid and M/13
phosphate of different pH. Spont. decol. time 115 min.

pH 6.63 6.74 7.00 7.18 7.82 7.60

Decol. lime 17 16 15 13 12 11 min.

The hexose-diphosphoric dehydrogenation is thus also sensitive to acid
reaction. The latter has rather little influence when decoloration is rapid,
but, if the spontaneous-decoloration time at alkaline reaction is already,
say, 80 min., only a slight reduction in the activity is required for de-
coloration to cease altogether.

In respect of the spontaneous donors as well as Ci ray ex-
periments yield results directly opposite those of Andersson’s.
With suboptimum quantities of Ci and strong spontaneous
activity my tests give a pH-curve that is very similar to that
Andersson found for the spontaneous donors together with
Ci. It is possible that in this test of his the spontaneous donors
played the dominant part and that his erroneous result is due
to this.

As, then, the spontaneous activity in the cucumber-seed ex-
tract is sensitive within a certain range to pH fluctuations,
it is necessary to work with buffered solutions. Buffering in
e.xtracts made with K2HPO4 ought to be sufficient in most
cases, but to be on the safe side I went over to extracting with
M/15 phosphate buffer of pH 7.38 — it may be suitable to bring
the hydrogen-ion-concentration within the range it usually
has in blood-serum \

^ In ray last e.xperimenfs, in which cucumber-seed belonging to another
delivery was used, it proved nccessarj' to extract with phosphate of about
pH 7.55 in order to get a suifabh* high spontaneous aclivitj*.

ON THE CITRIC ACID METABOLISM IN MAMMALS

Tlic possibilily of inaccurate values arising from a change produced in
tlic pH by added serum should a priori be negligible: firstl 3 ', the buffer-
effect of the serum itself is not so great (< 5 % of the whole blood’s)
sccondlj', the scrum is mostly considerably diluted in the fluids concerned.
And a pH-disp!acemcnl would make itself the more noticeable, the higher
up it occurred on the ascending leg of the curve but there the addition of
serum — and with it the possibilit 3 ' of a pH-displacement — becomes
progressivch’ less. If, moreover, the ascending leg is “in line’’ udth the
spontaneous-decoloration lime of the cnz 3 -me solution, there is in this, too,
an assurance that the added serum has not altered the pH. Nor have I
been able to ascertain any such at direct measurements.

The inaclivalion of heart muscle isocitricodehydrogenase by
phosphate ions, demonstrated by Adler et al. (1939) appears
not to apply to the cnz}'me system in ciicumber-seed extract.
An increase of the phosphate concentration certainl 5 ’^ gives a
longer decoloration-time, hut the retardation is of about the
same order of magnitude as follows an increase of the NaCl
concentration. Nor liave I been able to find any difference in
the intensity of the inhibition at different pH.

Experiment (15. 1. 40): Dccoloralion-timc at pH 7.7C with different

phosphate concentrations.

Ci in tubes

.... 2

10 •/

Phosphate M/80 . . . .

.... 54

14 min.

„ M/40 ....

16 „

Unless specially purified, aqueous extracts also contain some concen-
tration of phosphate, since the cucumber-seeds themselves contain rather
much phosphate; after the oil has been expressed, the residue contains
3.7 %, reckoned ns P-Oj (WEHMEn, 1931).

Adler ct al (1939) had also found that manganese or magne-
sium were necessary complements for isocitricodehydrogenase
action. If this is the case with isocitricodehydrogenase from
cucumber-seed, there are doubtless sufficient quantities of these ■
substances in the extract, for no activating effect follows their
addition in the concentration stated to be optimum.

Manganese (Mn sulfur. Kahlbaum p. analys.) was tested in concen-
tration from M/5000 to M/250 and had no effect. Magnesium [Mg sulfur,
puriss. cnjst. Merck] was indifferent (or possibly feebly activating) in the
low concentrations, but from MI500 it was inhibitive.

Experiment 5. 1. 40. Cucumber-seed e.xtract with M/20 KjHPO^. De-
coloration-lime with rising Mg:

JOHAN MARTENSSON

Ci in tubes

8 7

MgSO) 0

16 min.

—

la „

,, 1 ,a ,, . . • . .

.... 24

15 „

3.0 f, . . . . .

.... 31

18 „

j» 6.0 ,, „

26 „

Adler et al. (1939) also claim that monoiocloacetic acid is a
very powerful inactivator of heart muscle isocitricodehydro-
genase; M/lOO gives a 97 % retardation. This does not apply
to the isocitricodehv-drogenase in cucumber-seed extract. On
the contrary, monoiodoacetic acid has proved to be a suitable
addition for inhibiting the hexosediphosphoric dehydrogenase
without inactivating the Ci-dehydrogenase. 'Even so high a
concentration as M/40 produces only a 35 /o retardation.

Experiment 22. 1. 40;

In each tube 10 */ Ci

Monoiodoacetic acid . .

. . —

M/1000

M/200

Decoloration-time

.. 12

Influence of Light on Rate of Decoloration.

fhc works of Lehmann (1922), Krestownikoff (1927) and v. Euler
and Klussmann (1934) show that increased light quickens decolorization
in the Thunberg experiments. GrOnvall (1937) was tlie first to study the
influence of light in enzymic Ci-estimations. He found that if light was
kept from the tubes the decoloration-time was lengthened, while if
artificial light war added to the daylight the time was shortened. Tliis only
applies, however, if MB or thionine is used as indicator, and not indigo
trisulphonate, which is far less sensitive to light variations.

The great influence of light in Ci-estimating work is observed
without special tests. On a clear, sunny day decolorization
. is very rapid, on a cloudy, rainy day it may even happen that
tubes with a small quantity of Ci .never become decolorized.
.At times the light varies very much in strength while an ex-
periment is going on, which has a very disturbing effect. To
avoid this element of uncertainty I found it simplest to adopt
entirely artificial lighting: I could then retain MB as indicator,
which is easier to read off than indigo trisulphonate. Control
tests showed that the form of the curve is about the same in
electric light as in daylight of the same intensity. This applies
to both the standard curve and the X-curve, and hence the

ON THE CITRIC ACID METABOLISM IN MAMMALS 25

same values are obtained if the estimation is conducted in
strong light on one occasion and in less strong light on another.

Experiment 2S, 30 and 3J. 1. 1939. A series of estimations made in
(1) daylight, (2) electric light, (3) daylight plus electric light. To eliminate
effects from the change the enzyme is undergoing the whole time, the tests
were c.arried out on three days in alternating sequence as regards the
lighting. The series were then collocated, and the mean values of the
decoloration-times are given in Fig. 5.

Experiment 10. 1. 40. Electric light. Distance from source of light in
first series 100 cm, in second series 141 cm. The illumination is therefore
double as strong in the former case as in the latter.

Fig. 5. The citrate-curve for different intensities of illumination: I Day-
light, II Electric Light, III Day-Light + Electric Light.

Ci in tubes 0 2 3 4 5 6 7 8",'

I 71 45 31 20 15 13 12 12 min.

11. * 87 58 43 27 20 16 15 14 „

When decoloration is relatively rapid a change in illumina-
tion does not cause so great changes in the decoloration-time,
but if the latter is fairly long from the start a slight decrease
in the lighting may result in failure of complete decoloration.
A lengthening of the decoloration-time gives the inactivation of
enzyme or coenzyme going on during the experiment a better
opportunity of making itself felt.

The lighting arrangements were as follows; The estimations were
carried out in a room entirely screened from daylight. The water-bath
was illumined by two ordinary Osram lamps (lum. effect: 100 decalumens).

JOHAN mAkTENSSON

Theae were attached to a stand behind and above the person managing
the evacuations and readings, so that this work could proceed without the
tubes being shadowed; illumination from behind is also less straining to
the eyes. The distance from light-source to water-bath was at least
120 cm, the two lamps were 40 cm apart, while the series of tubes had a
maximum range of 60 cm. In this way differences in illumination in differ-
ent parts of the series can only be very slight and of no practical im-
portance. Bj increasing or decreasing the distance it is a simple matter
to vary the lighting intensitj' if decolorization is proceeding unsuitably fast
or slowly. When, as is usual, samples are estimated twice with different
enzyme preparations, they should be set up the second time in reverse
order to the first so as to eliminate anj' illuminating differences or other
irregularities in the routine work.

The form of the X-curve in relation to the Standard

Curve.

Reference is made below to some aspects of the X-curve
which seem in need of accentuation, not having received due
attention in the Ci literature.

The method is intended for work on very small Ci quantities,
so that disturbing substances in the test fluid can be neutralized
by dilution. But with rising dilution the liability to errors in
pipetting and reading-off is also increased. With the usual
tube-loadings (0.2 — 0.3 — 0.4 cc. etc.) the error liability for tube
0.2 cc, is two and a half times greater than for tube 0.5. In the
region in which the curve just bends round into the level-line,
an insignificant variation of the decoloration-time may pro-
duce very great changes in the values. For this reason it is
safest to rely on the values obtained immediately above the
bend of the curve, as recommended by Ostberg (1934) ; the
liability to error subsequentl}’^ increase the further up the curve
one comes.

When a Ci-rich solution is being estimated, it must be
diluted before the tubes are loaded, since in routine work one
cannot pipette up smaller quantities than O.i cc. with any pre-
cision, Otherwise it is obviouslj' of no consequence whether
the diluting is affected before or at the loading of the tubes,
provided use is made of the same diluting fluid (for instance,
0,2 cc. of an undiluted solution contains the same amount as
0.4 cc. of the same solution diluted 1:2). Excessive diluting,
however, wall restrict estimations to the upper, uncertain part
of the cun’e, which spreads out over several tubes. Obviously

ON THE CITRIC ACID METABOLISM IN MAMMALS 27

the usual type of curve cannot then be expected. Fig. 6 shows
the theoretical course of the curve for different dilutions of the
same Ci-solution. ^Vith a dilution of only 1; 4 the curve does
not extend down to the level-line. The curve 1 ; 4 implies that
the Ci quantities in the three top tubes of the 1: 2 curve have
been distributed over six tubes instead of three. Clearly the
relative differences between the tubes are smallest furthest
down the curve, and it may therefore be very difficult in prac-
tice to determine any difference in decoloration-time: one gets

Minutes.

too

Fig. 6. Form of tlic cilrate-curve for different dilutions of the fluid to

be tested.

a “false level” high up. The risk of this error occurring be-
comes greater the less Ci the solution contains; i. e. the longer
the decoloration-times become. With a long series of tubes
it may piay a minor part, but in routine work limited to three
or four tubes the error may be considerable if attention is not
directed to it. It is therefore safest to frg to get a dilution
of the test fluid that will give a curve as nearly as possible
identical with the standard curve.

It is evidently inattention to the conditions pointed out here which has
led to LENNfin’s (1934) statement that “the level is higher in estimations
of fluids with low Ci-content”, which was later confirmed by SKjsTRd^t

JOHAN mArTENSSON

(1937). It is then quite natural that he is able to stop the inhibilive action,
upon which the high level is thought to depend, by adding Ci in even small
amounts. No doubt the same misunderstanding is responsible for SjO-
stbOm’s reference to “sera -with a high level”, although he publishes at the
same time reports of an experiment on the same serum which in the 1 : 3
dilution had a much higher level, in the 1 : 2 dilution a considerably lower
one, although the inhibition there ought to be greater. In the higher
dilutions he could find no difference in decoloration-time for those serum-
quantities, which in the lower dilution gave a difference of 19 minutes.

From. Fig. 6 it is also to be seen that by the Thunberg me-
thod as customarily disposed (with 10 y MB in each tube) it
is possible with relatively high accuracy to estimate the Ci-con-
tent of solutions containing down to 5 7 of Ci per cc; for lower
concentrations the values are more uncertain.

The Systematic Deviation of the Serum-Oitrate Curve.

All who have occupied themselves with estimations of the
citric acid in the serum have found that the serum-curve is
often subject to a systematic deviation of the values, so that
these fall with rising amounts of serum. The ascending leg
of the curve is more horizontal and the level often somewhat
higher than in the standard curve. It may even happen that
the value for the 0.2 cc. tube is double as large as for the 0.6
cc. tube. Which of these values is the more correct? OsT-
BERG (1934) carried out tests with retarders and accelerators
added to known quantities of Ci and in both cases found the
least deviation from the right value in the lowest part of the
ascending leg of the curve.

Two alternatives must be borne in mind: 1) That serum
contains retarding substances whose action is the greatest in
the tubes containing most serum, i. e. where their concentra-
tion is the greatest. That this can actually be the case is readily
seen from the frequent failure of attempts to press the level
of the X-curve down to that of the standard curve, even bj'
reducing the dilution; instead, a longer decoloration-time may
even he got the more serum that is loaded into the tubes.
2) That serum contains accelerating substances whose action
is stronger the less Ci the tubes contain, that is to say, the
larger the proportion of MB decoloration dealt with by the
spontaneous activity. It seems very natural to suppose that
the most important cause of the systematic deviation is that

ON THE CITBIC ACID METABOLISM IN MAMMALS 29

scrum contuins, besides Ci, other substances that serve as do-
nors for dehydrogenases in the enzyme solution. It has been
previously shown that the spontaneous dehydrogenation de-
clines in intensity in the course of the experiment as a result
of the successive decrease in substrate concentration. When
this becomes very low, an extremely small addition of donor
substances suffices (when decoloration-time is relatively long)
to cause a distinctly accelerated decoloration. This effect be-
comes the more pronounced, and the values consequently the
higher, the further up the ascending leg of the curve they fall.

TJial lliese rising values aclu.ally depend on an inlensified acceleration,
and not on an intensified retardation, is evident from, e. g., the following
experiment in which a scries of tubes gave a particularly pronounced
systematic deviation of the values (serum diluted 1 :2):

Scrum quantity . . O.i 0.2 0.3 0.4 0.3 0.6 cc

Ci-valuc 37 25 21.7 19.5 17.C 15 y/cc

The 0.2 cc serum was replaced by 0.1 cc scrum -f O.i cc Ci-solution contain-
ing 1 of Ci. If the value for the O.i tube, 37 y, were correct, the tube
with 0.1 cc scrum + 0.1 cc Ci-solution ought to be found to contain
1.W+ 1 z=2.63 y, since this tube docs not contain scrum in a larger con-
centration than the O.i tube and consequently any retarding action cannot
be so great either. Actually it proved to contain about 2.3 y, and hence the
acceleration induced by oUicr donors was considerably less with this
shorter docoloration-timc.

It must also bo remembered that the sj-stemalic deviation may also be
due in p.art to the presence in the serum of co-enzyme, which, while not
affecting the Ci dehydrogenation, presumably does affect the spontaneous.
In favour of this argues the fact that the systematic deviation oftener
becomes more pronounced with fresh scrum than with serum that has
been kept for a day.

There is thus good reason to expect that the Ci-curve for
serum will he affected, in most cases, by an acceleration —
greater or less — that predominates in the upper part of the
curve and, in many cases, by a retardation that predominates
in the lower part of the curve. Consequently, the correctest
values are most likely to be found just above the bend of the
curve, where these two factors neutralize each other. But this
presupposes the use of a long series of tubes for the serum
Ci estimations in order to ensure that the point at which the
level-line bends round into the ascending leg and the proper
values really can be determined. Or else an attempt must be
made, after preliminary tests, to adjust the dilution so that

JOHAN MARTENSSON

the serum-curve as nearl}' as possible resembles the standard
curve, in which case only the three tubes which in tlie com-
plete curve fall immediately above the bend need be included
in the calculation. The latter alternative is distinctly more
practical and accurate. But if, in such a case, the dilution hap-
pens to be excessive (so that even in the top tube the decolora-
tion-time for the serum-curve is considerably higher than that
for the standard curve), this will mean that three values are
chosen high up on the complete curve, where the uncertainty
is per se considerable and where the systematic deviation as-
serts itself most, and hence the value obtained will mostly be
too high. That is why with apparently the same technique
widely different values are obtained from the same serum,
depending on the dilution selected.

Hirschlaff-Lindgren (1937) has estimated the Ci-contcnt of horse
serum, in •v\’hich work she also made use of a control curve of serum
solution with an addition of Ci-solution. She then comes to the conclusion
that the method only detects 90 % of the serum Ci. But in her experiments
an upper part of the curve is compared with a lower pari, and as the
systematic deviation of the values should suffice to explain the difference,
her conclusion that the method only accounts for 90 % of the serum Ci
is not justified on the basis of her experiments.

For routine estimations I proceed as follows. The standard
curve is adjusted by strength of enzyme and intensity of illu-'
mination so that the level lies at about 15! minutes and the
spontaneous decoloration at 80 — 100 minutes. So rapid a
decoloration is more advantageous from all points of view, e. g.
the estimation takes less time, the sources of error are fewer,
multiplying appreciably with lengthening decoloration-time.
The contact-point with 10 '/ of MB then comes to lie between
5 and 6 y of Ci. The enzyme solution can be adjusted to suit-
able quality by varjdng the dilution, the degree of stirring and
the extraction-time. After preliminary tests with 0.4 cc. of
serum solution against 0.4 cc. of standard solution, the serum
is so diluted that the solution may be expected to contain about
10 '/ per cc., and three tubes with respectively 0.5, 0,4 and 0.3 cc,
are used for the estimation. If with a dilution of, say, 1: 8, the
serum gives a value between 70 and 90 y per cc., there is
reason for beliewng this correct. Should there be consider-
able deviations, however, it may he advisable on repeating the
estimation to employ another dilution that gives better agree-

ON THE CITRIC ACID METABOLISM IN MAMMALS

ment with the standard curve. As the method is so delicate
that accidental factors may creep in, the same serum should as
a general rule be estimated on two separate occasions.

Calculation of the Oi-values.

For the working out of the Ci-values I have used Thunberg’s
graphic method, which is grounded on the basic principle thAt
under equal , conditions equal decoloration-times mean equal
quantities of Ci in the tubes.

LennBr’s method of calculation is more tedious and would seem to
give somewhat more uncertain values owing to the difficulty of exactly
determining the position of the contact-point and to the fact that the level
of the serum-curve frequently lies somewhat higher than that of the
standard curve. His statement that the contact-point values of the serum
and standard curves are equal even if the serum-curve level is iiigher is
theoretically inaccurate, although in most cases the assumption does not
lead to great errors in practice. It is correct provided the Ci-activity and
the spontaneous activity arc subject to an inhibition of equal degree, so
that the ratio between them remains unaltered. But this probably occurs
only in the “artificial serum” with which Lenn6r has sought to prove
the accuracy of his statement; in that serum there occurs only an inhibition
by salts, no substances are there that increase the spontaneous activity,

as there are in natural serum. ^

SjOstrOm’s method of computation is only a special case of Thunberg s,
applicable if the curve resembles a rectangular hyperbola: a mean value
for the serum-ciirve is compared with a mean value for the standard
whereas in Thunberg’s method the value of each serum-tube is «
computed and then the mean is taken. The advantage o J
method is considered to be that it renders possible a calculation o
curves that lie so high above the standard curve that they cannot be
evaluated direct. Curves of this type, however, give such unreliable resit
that they ought to be discarded from' the outset. No ^ ,

appear to be offered by SjOstrOms method as agams ^ hot
SjOstrOm has subjected Lennar’s method to a severe criticism, as
a series of experiments showing that Lenn^R’s method may lead W sue
absurd results as that a serum, after addition of Ci, ^ ,

than without this addition. But these experiments o^f SJOsxRdMS a
faulty owing to his having used serum in so great

not been able to determine the real ^Lh^d L foUhded.

point, just those characteristics upon which LennER

In his series SjOstrOm does not use so large quantities o s .

„ eve„ .heo.e.,c po^ibrn..

that this excessive criticism, based as it is

JOHAN mArTENSSON

have been instrumental in discrediting the enzymatic Ci-estimating method
as such.

The accuracy of the Enzymic Method of determining Oi.

By Thunberg’s method it is possible to determine quanti-
tatively the amounts of Ci added to serum or other biological
fluids, as is evidenced by many experimental investigations
(Lenner, Schersten, Sjostrom) and as I can confirm on the
strength of a number of experiments. There are, then, toler-
ably good grounds for concluding that even those values di-
rectly obtained by the method represent the actual Ci-content
of the fluid concerned. A more direct check has been devised
by Schersten (1936). He made a series of estimations on
genital-gland secretion simultaneously by Thunberg’s method
and the purely chemical pentabromacetone method, and found
no systematic difference between them.

Respecling the single estimation mean error the following figures are
available: Lennar (1934), on the basis of the differences obtained at
duplicate estimations of 19 sera, found the mean error to be 0.52 7 (which
represents about 2.5 %), From 31 duplicate estimations ScuEnsTiSN (1930)
calculated the relative mean error of a single estimation at 2.57 ± 0.83 % ;
he expresses the difference in percentage of the mean value of the two
estimations because equal relative errors are equally probable. SjOstrGm
(1937) computed a mean error that" under ordinary conditions should not
exceed ±0.8 %. But this figure seems more to be an expression of how
closely his curves approach a rectangular hyperbola, and it is debat.nble
whether the figure is a measure of the accuracy of the method. In any
case, his calculation onlj' partly covers the sources of error attaching to
the method. A reliable measure of the single estimation mean error can
presumably only be obtained by carrying out two estimations on a sufficient
number of sera on separate occasions and with different enzymes, so that
alt sources of error will have a chance of exercising their effect. From
such a scries of 28 duplicate estimations GrUnvall (1937) arrived at the
value 3.3 ± 0.44 %.

On comparing some duplicate estimations I found that the
dispersion could sometimes be still greater than is indicated by
the figures adduced above. For that reason I have since made
double estimations xvhen possible. A comparison of all the
150 duplicate estimations made during the period 1939 —
1940 gives the result submitted below: the dispersion is

calculated according to the formula

1 /2’D-

'V 2n ’

where n is the

ON THE CITRIC ACID METABOLISM IN MAMMALS

HumliBi of (ixiplictito ostimiitioDs jiDd 2) tli6 difference expressed
in % of the mean value of the t^YO related determinations,
since equal relative errors are equally probable.

Diff

No. of Duplicate
Estimations

180

224

300

684

490

768

729

200

847

432

169

225

n = 150

I’D' 5325

0 — 1^5325/300 = 4.21 ± 0.2i %

In routine work, therefore, the single estimation mean error
is 4.21 + 0.21 % . Assuming the accuracy of the value obtained
from a duplicate estimation to be V 2 greater than that of a
single estimation then the accuracj'^ would be very near 3 %.
Consequently, should two values, each determined by duplicate
estimations, show a difference of 12s %, this difference is to
be regarded as significant.

Another series of 200 duplicate estimations, most of which were carried
out during the spring (the whole series 1939), when according to

experiences the enzymic method is less accurate, gave by the same com-
putation a slightb' greater dispersion: 4.58 ± 0.23 %.

Comparative Investigations with the Pentabromacetone

Method.

The Thunberg method has certainly also been used for the
investigation of the Ci in tissues (Gemmill, 1934) , but, as this
can only be done after an extensive preliminary treatment, it is
no doubt more convenient and exact to use a purely chemical
method for this purpose.

JOHAN MARTENSSON

A review of Ihe various chemical mclhods for determining Ci has al-
ready been given by Scherst6n (1936). Tlie very sensitive colour reaction
described by FCrtii & Herrmann (1935) — not included in Scberst6n’s
review — has scarcely been elaborated yet for e.vact quantitative determin-
ation of Ci. In the present work I have used the pentabtomacetone method
as modified by Pucher, Sherman and Vickery (1936). It is founded on
tlie fact that Ci is oxidized to pentabromacetone, which with sodium
sulphide is transferred to a coloured substance, this being measured with
the Pulfrich photometer and the Ci-values thereupon obtained from a
calibration curve. By this method quantities between 0.1 — 1 mgm. can
be determined to an accuracy of ±5 %. Its sensitiveness and accuracy
are accordingly essentially lower than the Thunberg method’s, and the
dispositions described, which I have followed in detail, take up at least as
much time.

Besides estimations of the Ci-concentration in tissues and
whole blood I have made some experiments with the penta-
liromacetone method on serum for comparison with the Thun-
lierg method. The values obtained, however, are not directly
comparable, since the chemical reaction is given exclusivelj"
by Ci, and not by cis-aconilic acid and isocitric acid (Breusch,
1037), which are in some degree of equilibrium with Ci in
the tissues and which are also included in estimations with
the enzymic method. Statements as to the proportions in
this equilibrium differ somewhat: at least 75 % Ci (Breusch,
1937); 90 % Ci, 10 % isocitric acid, and minimum concentra-
tion of aconitic acid (Martius, 1938); 80 % Ci, 16 % isocitric
acid, and 4 % aconitic acid (Johnson, 1939). No investigation
is available as to the conditions in the serum, and, of course,
nothing is known of the distribution in blood from different
parts of the body. Large series of parallel estimations with
the chemical and enzymic methods would be required to in-
vestigate this interesting problem and the question of the trans-
formation of orally or parenterally administered Ci, since both
the methods are burdened with an error percentage that is ra-
ther large in proportion to the differences which are likely to he
found. In view of this I have not gone into the question here,
my object being to ascertain whether biological experiments give
the same result when Ci is estimated with the purely chemical
method as when it is estimated with the enzymic.

In a series of 13 estimations on serum from arterial blood the
pentabromacetone method gave values which were on an aver-
age 16.5 % (10.5 + 2.9 %) lower than those given by the Thun-
])erg method.

ON THE CITRIC ACID METABOLISM IN MAMMALS

The degree of accuracy I attained with the pentabromacetone
method has not been determined on the strength of any large
series. A comparison of nine duplicate estimations gives the
single cslimalion mean error, calculated as described above for
Thunberg’s method, as 9 . 5 g + 2.23 ^ .

Summary.

The complex dehydrogenating process that underlies Thhn-
berg’s enzymatic method of Ci estimation is discussed. My
investigations have shown that:

The specific enzyme system in the cucumber-seed extract
for Ci shows good stability, even at a temperature of 35° C.
Tl)e spontaneous dehydrogenating activity in the enzyme solu-
tion, on the other hand, rapidlj' declines owing to exhaustion
of donor substances. The cucumber-seed used contains itself
about 0.035 % of Ci, which must be dehydrogenated before tlie
typical citrate-curve can be got by adding Ci.

The specific enzyme system in the cucumber-seed extract
for malic acid is less stable than that for Ci and for hexose-
diphosphoric acid.

The activity of the Ci-specific enzyme sj’stem in the cu-
cumber-seed varies but slightly, if at all, in response to changes
in the hydrogen-ion concentration within the range pH 6 . 6 — 7 . 6 ,
whereas the spontaneous dehydrogenation declines with falling
pH and ceases almost altogether at pH 7. The dehydrogenation
of malic acid and hexose-diphosphoric acid declines with falling
pH.

The most important cause of the systematic deviation in the
vaUie.s of a serum citrate curve is doubtless that serum, in
addition to Ci, contains quantities of substances which serve
as substrates for dehydrogenases in the enzyme solution.

For the practical performance of the Ci estimation the fol-
lowing points have proved to be of importance:

The stirring of the cucumber-seed extract must alwaj's be of
equal degree if equal enzyme strength is wanted in the extracts.

If the spontaneous dehydrogenation in the extract is too
strong, a suitable enzyme solution is not obtained by diluting,
but by keeping the solution in room temperature for a while.

The lighting should be kept constant, preferably by means
of artificial lights in a room screened from da 3 'Iight.

The dilution of the solution to be tested should be so selected

JOHAN xMARTENSSON

that the X-curve follows as closely as possible the course of
the standard curve. The most accurate values lie immediately
above the bend of the curve. In routine work the single
estimation mean error proved to be .4.21 + 0.24 % .

The pentabromacetone method in Pucher, Sherman and
Vickery’s modification, which is specific for Ci, generally
gives lower values for serum than the Thunberg method, which
also includes any preformed isocitric acid and cis-aconitic acid
that may he present.

CHAPTER II.

The Citric Acid Metabolism in the normal
Manunalian Body.

Review of previous investigations.

O n the strength of his investigations into the effects of certain vegetable
acids, among them citric acid (Ci), on the elementary respiration in
minced frog muscle, Thunberg (1911 a, b, c) surmised that Ci is a normal
intermediate product in the metabolism of muscle. Batelli & Stern
(1911) found that the consumption of oxygen was also increased by Ci
in the liver, kidney and muscle of mammals. Tlic specific dehydrogenase
for Ci was subsequently shown to be present in most tissues. Wishart
(192.^) demonstrated citricodchydrogenase in phosphate extract of liver,
and Bernheim (1928) submitted a method of extracting Ci-dehydrogenase
from liver, which method has since been improved by ScherstEn (Wd )
and SjOstrOm (1937). As a result the liver-cells have been shown o
contain Ci-dehydrogenase in abundance. Using a modification o ^
pentabromacctone method, Langecker (1934) measured the brea ' o
Ci in various minced tis.sues from the rabbit. Her experiments sbow-ed
that the liver possesses great ability to destroy Ci, the kidney
less, and muscle very little. Ci added to intestinal contents was es
in only a slight degree, and hence bacterial decomposition of Gi m U
intestinal lube is not of any importance. Nor was Ci “yed in
blood, a fact that had already been shorvn by Salant & i ( '

These authors were the first to use a toler.ably
study of the in vivo Ci metabolism of the mammalian body ^

modification of the DenigEs method

in blood samples from rabbit, cat, and dog. npriments were

Ci o„IIy or .ubciooooosly (In ooch l»Bo do«s IMl *''' “ “

mo,o or loxicoloeicol intor.st), n consldernble nmomt

co,orod In Iho blood. Ador Inlmvonons inieclion tho C. ^n.ohod qo.Aly

from the blood. Only a small proporlion TOS tonnd i .

atlcr snbculancous injocllon, In rabbil nrlne on an aroraee

30 %, aUbouBl, Ibo dinrollo etfool n-as

Tho first qnanmallvo invcsliBalion of the C normally

c.arricd oat by OsTBB.m (1931). Tho 24.hour amonnl of C, vane, norm.

JOHAN MARTENSSON

between 0.2 and 1 gram. Intake of Ci by mouth, even in so large a dose as
40 grams, does not appreciably increase the Ci excretion. Practically all
the Ci-intake is therefore oxidized in the body. This result was soon
confirmed by other investigators.

Kuyper & Mattill (1933) studied the Gi-content of blood from differ-
ent parts of the body in rabbits, but could not find definite evidence of
where in the body the Ci was formed or oxidized; the renal-vein blood,
however, contained less Ci than the arterial blood. Protracted inanition
lowers the serum citric acid (C/s, given hereafter in 7 = micrograms per
1 cc.), but even then Ci goes on being excreted, which indicates that it is
endogenously formed. (It was earlier known that Ci is introduced with
the food, since a large number of foods contain Ci; see Ostberg (1931) for
the literature on this subject.) Boothby & Adams (1934) found the same
in the case of the dog, and it was demonstrated still more convincingly
by Sherman, Mendel & Smith (1936 a). They followed the Ci excretion
in dogs during a long period of experimental alkalosis, when the 24-hour
amount of Ci in the urine amounted to 100 — 300 mgm. against 1.3 — 8 mgm.
normally. During three stated test-periods the Ci excretion was respect-
ively 5, 10 and 7 grams larger than the Ci intake with the food, which
clearly shows that Ci is a product of metabolism. For there is no store
of preformed Ci in the tissues, the Ci-content there being of about the
same order as that of whole blood, as Pucher, Sherman & Vickery (1936)
have shown with their modification of the pcntabromacctonc method *. In
liver, however, they found very little Ci. (Still earlier Gemnhll (1934)
had been able to show the presence of Ci in the tissues, while Langecker
(1934) stated that normally- the tissues do not contain Ci.) Tlie dog
excreted with the faeces normally 0.4 — 0.8 mgm per day, and this quantity
was increased very slightly after oral administration of Ci. Sherman,
Mendel and Smith (1936 b) studied the C/s in dogs after oral ad-
ministration of Ci: after 1 gram of Ci per kg of body weight the C/s in-
creased to 2 — 4 times the normal value, reaching maximum after — 3 ‘/s
hours and returning to normal within 3 ‘/s — 7 ^/j hours. Simultaneous
estimations of Ci in blood and urine appeared to point to the occurrence
of a renal threshold for Ci in the dog, though this threshold value varied
widely in different individuals.

Still earlier, the C/s had been investigated after oral Ci administration
in tolerance tests on man. Thunberg (1933) found, in a normal Subject,
a moderate rise (38.4 % after 15 grams of Ci) that lasted some hours, and,
in a patient with lesion of the liver parenchyma, a more powerful and
more prolonged rise. Even 5 — 8 grams of Ci by mouth give a distinct
elevation of the C/s, with maximum after 1 — 3 hours (LENNfeR, 1934). In
his comprehensive and important work on the C/s in hepatic diseases
SiOsTRtiM (1937) reports that 1 — 2 grams of Ci by mouth scarcely raise
the C/s in normal persons, while such a dose may produce a distinct and
lengthy rise in patients with a lesioned liver parenchyma. Tolerance tests

‘ Cr. footnote p. 65.

ON THE CITIHC ACID METABOLISM IN MAMMALS 39

with Ci given orally can Iherefore be a valuable supplemenlation of the
livcr-funclion lest that estimulion of the C/s constitutes. After intravenous
injection of O.ri grain of Ci the C/s conditions arc substantially the same;
in normal persons a moderate rise, usually with a return to the normal
value in BO minutes; in abnormal hepatic function with a C/s already high
n powerful and protracted rise. GiiOnvaix (1937) followed the C/s in
rabbits at relatively .short intervals after slow intravenous injection of
fairly large amounts of Ci: the transient rise was slight as against the Ci
quantity supplied, and hence these Ic.sts, too, show that the body has
great possibilities of coping with Ci introduced into if.

In connexion with his clinical investigation SjOstrOm (1937) also made
perfusion test.s on isolated liver from rabbit, cal, and dog. He found that
the liver has an enormous power of metaholizing Ci: even when the
perfusion fluid contained 1(10 limes the normal amount of Ci, this was
restored to the normal value by a single passage of the perfusion fluid
through the liver. Thus the liver acted as an efficient Ci filter, up to
20 grams of .sodium citrate being removed with ease from the perfusion
fluid hr n rabbit liver. On perfusion of muscle preparation he obtained
no corresponding consumption of Ci. In experimental lesion of the liver
induced in rabbits with allyl fonniate, the C/s graduallj’ rose, and on
perfusion a liver lesioned in this way proved less able than normally to
metabolize Ci.

The knowledge we pos.se.ssed of the Ci metabolism of the
body at the time the present work was commenced may be
.sumniarized sis follows: Ci is taken into the body with the food
but is also formed endogenously. The general occurrence of
Ci in all body fluids suggests that this formation takes place
fairly generaily. The body possesses a very great power to
break rapidly down inlaken Ci, only a very small proportimi-
being excreted in the urine and faeces. This breakdown o i
is mainly effected in the liver, which has proved to contain
an abundance of Ci-dehydrogenase and which at perfusion
tests has shown an enormous Ci-metabolizing capacity.

The Citric- Acid Content of the Serum after Intravenous
Administration of Citrate.

My o^yn cxpcrlmcnls started with ,

on rabbits and cats. By following the C/s ^

intervals after a one-time intravenous dose of Ci " ““ “

get a rough idea of the extent of the C. breakdown o a
Suitable dose that docs not elevate the C/s fo abo^ m

norm.aI concentration. Fiff. 7 rvill. text embodies some typical

JOHAN MAKTENSSON

tolerance tests, which have been confirmed by many similar
ones.

For the experiments I used the preparation Sodium citrate Analar,
British Drug Houses. Tlie dosage is given in corresponding amounts of
citric acid (with 1 mol. HbO), since as standard both for the Thunherg
and pcntabromacelone methods I used Acidum citricum cryst. pro analyst
Schcring-Kahlbaum, which contains 1 mol. HsO (which is possibly inlra-

C/s -x/cc.

Min.

Fig. 7. CIs after intravenous administration of Ci.

1. Cat, 3.5 kg. Pernocton 0.7 cc/kg. 43 mg Ciihg.

II. Rabbit, 2.1 kg. Urethane 1.25 gm/kg. GO mg CilFg.

III. Cat, 2.4 kg. Urethane 1.2') gm/kg. 100 mg Cilkg.

IV. Ralihit, 2.5 kg. Not narcotized. SG mg Cilkg.

V. Rabbit, 3.9 kg. Urethane 1.25 gm/kg. 100 mg dikg.

molecularly bound, Thunberg 1933). By intravenous administration one
avoids absorption inequalities and can reckon with an exact quantity. The
speed of injection, which is entirely decisive of the toxiciti' of intra-
vcnouslj' introduced Ci, was kept between 1 — 2 cc. per minute (according
to size of animal) of the isotonic 3 % solution: in doses up to the higliest
u.sed here, 100 mgm. per kg., it docs not produce any pronounced reaction.
The pharmacological effects of Ci arc not dealt with in this work, and

ON THE CITRIC ACID METABOLISM IN MAMMALS

were not recorded. The blood samples for Ci-esliniation (3—4 cc.) were
drawn from the A. feraoralis or carolis of the narcotized animals, from
the auricular vein of non-narcolized rabbits. As narcotic I used in most
cases urethane in relatively small dose, 1— 1.25 gm. per kg., subcutaneously,
re-inforced with ether for operations, in some cases pernocton.

TJie experiments confirm earlier observations of the body's
great power to break down CL This makes it possible to study
the metabolism in fairly short experiments, which very much
facilitates the experimental analysis. The curv'es show that
a disappears from the blood the more rapidly the higher the
concentration is. This rapid fall at high concentration cannot
be due solely to a possible levelling-out of the concentration in
the tissues just after the injection, for it is also found at pauses
in long continuous infusion, when any such levelling-out ought
already to have occurred. In one such experiment on a dog
the C/s fell in 90 seconds from 925 to 700 7 per cc, in an-
other from 685 to 565 y per cc. In an experiment on a rabbit,
in which the C/s had been forced up to 1540 7 per cc., it fell
by over 400 7 in 70 seconds.

A significant point is that the breakdown of Ci in the nar-
cotized animals seems to proceed at the same rate as in the non-
narcotized, which suggests with a rather high degree of pro-
bability that other results gained from experiments on nar-
cotized animals are also valid for the normal Ci metabolism.

Salant & Wise .(1910) found, as was mentioned before, that the urinary
Ci excretion after subcutaneous administration was larger in the cat than
in the rabbit, and from this they concluded that the rate of oxidation of
Ci is considerably quicker in the rabbit, as being a herbivore. This logical
conclusion received no confirmation on actual measurement in my ex
periments. On the contrarj', the curves show that Ci is eliminated more
rapidly in the cat, and therefore the greater Ci excretion cannot be due to
the C/s remaining high for any considerable length of time.

The Citric-Acid Content of the Red Blood Corpuscles.

In order to compute the Ci metabolism in certain perfusion
experiments it is necessary to have some conception of the 01 -
content of the red blood cells, since the Ci estimations are
carried out on serum or plasma but the fluid volumes have
reference to blood. Some experiments relating to this question

are therefore inserted here.

JOHAN MARTENSSON

NonofiO and ScnEnsTfeN (1931), using the pentrabromaccfone method,
estimated the Ci-content of the red blood corpuscles in a case in which the
plasma contained 35 '/ per cc. Tliey found about 10 y per cc., a part of
which, however, must be referred to plasma left on the cells. PucHER,
Sherman & Vickery (1936) determined the concentration in the total
blood and plasma and from that reckoned the amount in the corpuscles.
They found considerably higher values: the plasma/corpuscle quotient
was about l.c.

I have made some indirect determinations by the Thunberg
method, adding a weighed quantity of Ci to 100 or 200 cc. of
heparinated idood, which was then allowed to stand at least
12 hours. The C/s was thereupon determined. The corpuscular
volume had been determined with a haematocrit and the C/s
had also been estimated before addition of Ci. With these data
I could fix how much of the added Ci was present in the plasma,
and from this estimate the amount in the corpuscles. Later
on, some experiments with the pentabromacetone method were
made to determine the Ci-content of whole blood and plasma,
and the amount in the red corpuscles was calculated from this.
The results are given in the following tables.

Thunberg Method;

Plasm.!

Ci y/cc

Bl. Corpuscles
Ci y/cc

Plasma

Vol X

Plasma/Cclls

quotient

Ci in Blood
in ^ of C/s

Rabbit 278

70.0

62.5

4.0

305

60.8

57.0

5.0

))

248

60.0

74.5

4,1

» 313

37.0

59.0

8.5

Dog

270

70. 0

65.0

3.9

» i8ri

22.0

48.0

8.4

Pentabromacetone Method:

Plasma

Ci y/cc

Total Blood
Ci y/cc

Plasma

Vol ^

Plasma/Cells

quotient

Ci in Blood
in of C/s

Dog- 57.0

33.0

52.0

9.0

» 34.5

17.3

43.0

8.0

Cat 428.0

325.0

64.0

3.0

» 43.5

32.C

62.0

3.0

Rabbit 169.0

95.0

52.0

11.4

» 1140

770

62.5

6.7

ON THE CITRIC ACID METABOLISM IN MAMMALS

Of neither procedure can any great accuracy be expected.
The values, in fact, show rather wide deviations, but in no case
did I find such high values as Pucher and others. The con-
centration in the corpuscles is probably five to six times lower
than in the plasma.

When measuring the Ci metabolism I took as an approximate
value for the ratio of whole blood to serum the percentage
obtained if the serum-volume percentage is increased by about
Vjo. For instance, if according to the haematocrit reading the
serum volume is 65 % , the Ci-content of whole blood is assumed
to be 65 + 6.5 = about 72 % of C/s. This value probably comes
rather near the correct one (see the tables).

The Citric Acid Metabolism in Perfused Isolated Liver.

Even if the breakdown of Ci after intravenous administration
is verv rapid, it still takes at least one hour for the C/s to ^
to Us initial value. This is longer than

liver immediately oxidizes the Ci in the blood Howing through
it, as SjOstbO.M (1937) found in his perfusion experiments,
therefore repeated these experiments on rabbits and cats.

I used the perfusion .ppemlus .nd foU.ued .he
emptoycd .1 .he Luudspuerd Ins.i.u.e

1936), the foremost advantage of vfhich is that th method

liver'need not be interrupted more than at ^ T^ve

was last described by Blixenkrone-Moller (1938). '
bad an opportunity of studying the technique. ^

avoided where enzymic Ci estimations are undiluted

fibrinate the blood, but used heparinated possible.

in order to approximate iU reLtion stabilized by bubbling

The blood was fully oxygenated and its re^

0,-f 5 % CO. through it. A period of 15 20^ “ ^ .vas

balance and stabilize the conditions lonced even if technical con-

commenced. The latter was not unduly j^oUsm can scarcely

siderations permitted this, for a tolerably Ci .-as either

be expected to last more than ^ relation to the flow

added as a one-time bulk dose ® to spread uniformly over

volume per minute that it cou •nf„sed continuously over a long

the entire volume of blood) or was i experiment

period. The volume of blood was meas haemotocrit value

and the Ci-content was calculate wi , approximate so far as

us alruody dccrihed. This tf “

absolute values are concerned. What

JOHAN MARTENSSON

conclusions can be drawn from the relative values as to the metabolism
under different experimental conditions, if the calculations for the differ-
ent periods of the experiment are carried out along the same lines.

The curves in Fig. 8 show the Ci-content of the perfusion
blood after addition of a one-time dose of Ci. I got curves
agreeing with these from some ten other experiments in dif-
ferent modifications, even on perfusion with Tyrode’s solution

c/s y/cc

Fig. 8. The Ci-metabolism in perfused liver.

I. Rabbit, 3.7 kg. Urethane 1 gm/kg. Wt. of liver 128 gm. Perfusion fluid
220 cc heparinated blood (with aht. 20 % Tyrode's sol.). Addition
90 mg Ci.

II. Cat, 3.2 kg. Urethane 1 gtn/kg. Wt. of liver 52 gm. Perfusion fluid
170 cc heparinated blood (with aht. 40 % Tyrode’s sol.). Addition
GO mg Ci.

alone. The experiments show that the liver is by no means able
to effect on immediate elimination of Ci from the blood; on
the contrary, the isolated liver has little power to metabolize
Ci. In the rabbit’s liver the Ci metabolization is very low, in
the cat’s liver somewhat higher. But here, too, it is not suf-
ficient to explain the body’s power to break down Ci, even
under the assumption that the metabolic processes are more
active in the in situ organ than in the isolated. (As regards
the consumption of O- there is a decline of this in the perfused

ON THE CITRIC ACID METABOLISM IN MAMMALS 45

liver, by 25--35 To during a 2-hour period, Blixenkrone-
Moller, 1938).

In the following experiment with continuous administration
of Ci, in which the C/s did not rise noticeably above the level
normal for rabbits, the consumption of Ci can be estimated at
1.3 '/ per minute per gram liver.

Experiment IS. 11, 39. Rabbit, 3.4 kg. Urethane 1 gram per kg. Wt.
of li\cr 120 grams. Perfusion fluid: 190 cc. heparinated blood with about
30 % Tyrodc’s solution; hacmatocrit value 27 %. Min. vol. 48 cc. Pressure
12 cm, HjO. Temp. 37.8°. From 11.55 a. m. to 12,30 p. tn. addition in uniform
drip of 10 cc. 0.23 % sodium citrate in Tyrode’s sol. = 20 mgm. Ci. C/s in
perfusion blood (sample from stock tube): 11.55 a. m. 54.4 y, 12.15 p. m.
90.2 y, 12.35 p. m. 133 7, all per cc., thus a rise of 79 ’/ per cc. during the
period of 40 minutes. Tlie increase of Ci in the liver and the blood was
calculated as follows: To balance the Ci in the liver and in the blood left
there, 30 cc., corresponding to one-fourth of the weight of liver, were added
to the blood volume, which was 190 cc. Tlic Ci-content of the blood was
estimated at 80 % of the serum content, j. e. the increase was reckoned as
80 9o of 79 which is 13.0 mgm. on the total volume of 220 cc. Sub-
tracting this from the added 20 mgm., we gel a Ci consumption of 6.1 mgm.,
i. e. 1.3 / per minute per gram liver.

For such an immense breakdown of Ci as SjOstbOm obtained in his
experiments I have been unable to find a definite explanation. His ex-
perimental method, however, has certain weaknesses, more especially the
absence of effective arrangements for maintaining a constant temperature.
The explanation may lie in lliis, especially as SjOstbOm points out that
normal liver temperature is an imperative condition for obtaining this
great conversion of Ci. Furthermore, SjOstbOm assumes that the liver
consumption of Oj is verj’ moderate, and that a supply of oxygen from
(he portal blood cannot he c.xpected. The fact is, however, that the con-
sumption of Oi by the liver tissue is very large; in the cat it is normally
about 2.5 cc. per min., i. c. about ten limes larger than in a corresponding
amount of muscle (Blixenkbone-Molleb, 1937). According to Me Michael
(1937) the liver of the cal gels about two-thirds of its oxygen from the
portal vein, while the liver of the rabbit draws almost all its oxygen from
the hepatic artery. Tlie cat’s liver can take up all oxygen in the blood
flowing through it, but in the rabbit the 6:-content of the hepatic-vem
blood never falls to such low values. Now, since the circulation through
the hepatic artery is not maintained during the perfusion, the larger must
be the Oj-content of the blood delivered by the portal vein. It is esndent
that the liver used in SiOstkUm’s experiments was in some way injured,
so that the Ci and the Ci-dehydrogenase so abundantly present in the iver
had come into contact in a way that does not occur under physiological
conditions. On one occasion I obtained a considerable breakdown of Ci on

JOHAN MARTENSSON

perfusion of a rabbit’s liver, but in that case, in the course of the ex-
periment, the stomach contents had digested the stomach wall and lower
surface of the liver over a large area.

As control, with the liver in situ, 1 made an injection of Ci
in a branch of the portal vein. This produced a vigorous rise
of the C/s even in the arterial blood.

Of four concordant tests I am submitting the following: 10. 1. 38. Cat,
3.3 kg. Urethane narcosis. Ci slowly injected into the splenic vein in a
dose of 32 mgm. per kg. Sample from carotid arler}'.

Before inj. 7 20 40 60 min. after inj.

C/s 45.3 187 111 80.6 66.9 y per cc.

Blood specimens drawn at the same time from the portal
and hepatic veins did not, either, show so great a difference
in the Ci-content as to suggest any considerable Ci break-
down in the liver. It must be observed that the hepatic vein
also takes up blood from the hepatic artery, which always has
(as is shown later) a lower Ci-content than portal blood. This
agrees with a couple of experiments on rabbits by Kuyper &
Mattill (1933).

The samples used in my experiments were drown in the following
manner: The portal vein was laid bare, as was also the V. cava inf, in the
thorax, which was opened under artificial respiration. Tlie V. cava inf.
was ligated immediately below’ the liver, whereupon the samples were
withdrawn in rapid succession by puncture of the V. porta and of the
thoracic part of the V. cava inf. The C/s of the hepatic-vein blood was in
all cases lower and the differences, expressed in percentages of the portal
value, were, for four rabbits, 7.8 %, 2 %, 9.4 % and 7.1 %, for a cat, 2.c %.

After this it was easy to see that so vigorous a Ci metabolism
in tlie liver as that found by SjOstrOm was already excluded
by the fact established by several authors that the C/s can be
raised by oral administration of moderate doses of Ci, this even
in normal persons or experimental animals without any Eck-
fistula mechanism.

The Citric Acid Absorption from the Intestine.

Since a certain amount of Ci is ingested in the food, it is but
natural that one should find a higher Ci concentration in the
portal than in the arterial blood. After oral intake of Ci the
C/.S in man and dog reaches, according to works previously
cited, a maximum within ^/j — 3 Vs hours and returns to the nor-

o.v THK crnctc ach) mktahohsm in mammals

m:tl Mthte itt u f«nv hutirs. T/u; same seems (o apply to the
rahtut :u,Tor<liin' to tfie folioNviu" experiment:

V;,lu<; L'l 2 -I Jio(trs:iflcrC{-

nflniinlslmlioii

J -'ht' 70.r. r,7.o iVj.o y per cc.

•' It ...... Bl.j 7I.H m.r, y „ ,,

U5/( rijirritftcnt^ Ihc CV wns follnwcil in tlic porinl l)loocl nflcr in-
jrcliofi rif '.rtrlittnt filr.-iic ittli) Oic Tlu' liiouii spi-ciinens were

lal-rn ffrttti fi 1 -ennmiLi in tin- iiii-senffric vein. In n rabijif,

wtiich r,i~i \sr<i !Of» rnfan. Li per If;., the Cla rose in ■{;'> tnimilos from
jI*. htilicil f.tlnr- of }02 y (n ].'>/ jier rc.; in n c.'il, wliirli had hecn pis’cn
LO fiicni. f.i per Lp,, it ro'c from oOri ;• to TSi.s */ per rc. in .W niimiles.

I Inis the aiisfirption hefiins very rapidly, and according to
ptevioiis tests it romes In lui end in a few honr.s. Therefore
it is natnrnl that the (.’ s slunild fall during a fasting period.
I.LS'DiiOLM (tPiH) fonnd in rahhils on mixed rations a fall from
ahont Jf)p 1(1 (),') y nflrr 2t-hoiir.s fasting. I have followed
the C/s in a eonph* of c:ises during a -J-dav j)eriod, dtiring which

iter.

3 4 days in.anition

TH.n fit. I 05.0 {)('r cc.

fia.o .55,0 51. ■] y „ „

'rile decline of the CJs aj»pears tn he mo.st rapid during the
fir.st 21 honrs, tlumgti it also conlinncs afterwards. This fall,
linwever, need not only he due to inicrrnpled snjiply of Ci, but
may he caused indirectly hy onset of acido.sis due to .starvation.
In jiracfir.'dly all ex{)er>nient.s in lliis woi'k the animals had
fastf'd at ((•.•i.s! 2 } hours before being used. Then the sjiontaneous
faJ! of the C/s hikes pince so .slowly tlial it cannot possibly
infhience the ro.snlt of the, gcncraUy, short experiment^.

A higher concenlrntion of Ci is however also found in the
portal blood of experimental animals that have fasted up to
72 honr.s (,se table below). In view of the very ready absorb-
ability of (a shown in tolerance tests, it is scarcelj' credible
that any of the Ci that is preformed in the food can be left
in tiie intestine all this lime. Rather must it he supposed that
this Ci is a product of bacterial processes in the intestine or of
(he metabolic processes taking place in the intestinal mucosa

the rabbits were given o/dy w;

.Vurmnt ,

hntitnt t IfC) f.S.r.

It 7.s,r' ns.:-

JOHAN MARTENSSON

during absorption of the food. Thus it would seem that even
under fasting conditions the intestine is continuously delivering
a to the blood flowing through it.

Animal

Starvation

Days

Arterial Blood
C/s y/cc

Portal Blood
C/s y/cc

Diff.

of arterial C/s

Rabbit

42.7

53.t

25.5

))

42..’i

60..S

42.4

66. c

80.S

20.4

62.0

79.4

28.1

57.4 •)

lOl.o >)

75.C

Cat

61.0

79.0

17.9

53.3

72.7

36.4

))

67.8

80.3

18.4

57.8

72.9

26.1

^ By the pentabromacelonc method.

In the rabbit the large part of the coecum does not evacuate
its contents after several days’ starvation. That a bacterial
production of Ci can play some part there is shown by the
following experiments.

The contents of the coecum of a rabbit just killed were withdrawn,
mixed, and divided into two portions. To one of these trichloracetic acid
was immediately added, while the other was allowed to stand for two
hours at 39° before addition of trichloracetic acid. The Ci-conlcnt was
then determined by the pentabromacctonc method. Of two rabbits which
had fasted three days, one showed a' concentration of Ci in the intestinal
contents before and after incubation of 30 and 53 y per gram, the other
49 and 04 y per gram. In a third rabbit, which had been starved 30 hours,
the Ci-content rose from 54 to 79 y per gram. In this case the contents
left in the s.tomach were also analysed and found to contain only 13 y Ci
per gram.

Citric Acid Content of Serum after Functional Elimina-
tion of the Liver and Portal Area.

In view of the rather inconsiderable oxidation of Ci that
takes place in the liver, it may be expected that the Ci meta-
bolism of the body will not undergo any notable change if the
liver is put out of function. This was investigated by the
following simple means.

ON THE CITRIC ACID METABOLISM IN MAMMALS

The iirtenal branches to the portal region (coeliac art., sup. and inf.
mesenteric) were first tied and after that the portal vein. Then the gastro-
intestinal canal could be removed after double ligation of the oesophagus
and rectum. The liver was allowed to remain, as it cannot be totally
removed without injuring the V. cava inf. It was accordingly still in

c/s' y/cc

Fig. 9. Cis after functionaUij cutting off liver and portal area.

I. Rabbit, 3.1 kg. Urethane 1.5 gm/kg. Ligation of arteries to portal area
and of portal vein.

II Cat 3 7 kc. Pernocton 0.7 cc/kg. Same operation. . . „f

III. Rabbit, 4 kg. Urethane I gm/kg. Same operation. Intrav. mjec ion

60 mg Cilkg.

■.p.„ive- con.n,„„icali.n ,vi,h .he d,c„l..io». “ “ds

and outflow Of blood via the hepatic veins arising at P J

plays no par. for .he purpose of u,

“ .r as’’"”™? aTu.! died in a fe« hours, even .honph

JOHAN mArTENSSON

the decrease in the blood sugar resulling from the cutting-off of the liver
(Mann and Magath, 1924) was counteracted hy intravenous injection of
glucose. For the rabbit this requires, according to Svedberg, Maddock
and Dhury (1938), about 150 mgm. per kg. per hour, and in itself this
amount has no effect on the C/s. On the other hand, after a rather large
one-time dose of glucose (about 0.7 gram per kg. b. w.) I have found a
distinct fall of the C/s in a couple of cases.

Three characteristic cases from these experiments are shown
in Fig. 9. It will he seen that after functional elimination of
the liver and portal region the C/s of the rabbit rapidly falls and
subsequently remains at a low level. This is evidently due to
the fact that the flow of Ci (prevented here) from the portal
area is much greater than the liver’s power of breaking down
Ci. In the cat these two quantities are evidently better balanced,
the C/s altering but little. Intravenously administered Ci is
eliminated from the blood of the rabbit at about the same rate
as in an intact animal, at all events to begin with. This clearly
shows that the liver is not necessary for the breaking down of
Ci in the body.

As regards the method it may he objected that these "acute” e.xperiments
involve such shock effects from the operation that no conclusions can he
drawn respecting the metabolism. But the e.xperimcntal conditions are
kept constant to some extent hy the state of narcosis under which the
animals He during the entire experiment, before, during and after the
operation. Then, too, in experiments involving operation in several stages,
c. g. according to Drury (1929) or the still more sparing one according to
Himsworth (1938), the pure effect of bye-passing the liver does not
appear, as the animals have had ample lime to adapt themselves to such
changes in hepatic function as are implied in the diversion of the portal
blood direct into the general circulation.

The Citric Acid Metabolism in an Eviscerated Prepa-
ration and in Perfused Muscle.

Previous experiments have shown that Ci can be oxidized in
a preparation in which the kidneys and muscle play the great-
est metabolic part. If, in addition, the renal vessels are tied,
giving a preparation of predominantly muscle, there is a grad-
ual but considerable rise in the C/s of both rabbits and cats
(see Fig. 10: 1 and II), which cannot be interpreted otherwise
than that Ci is formed in the muscles No undoubted Ci-oxi-
dation follows an intravenous supply of Ci to an eviscerated

* Cf. footnote p. 05.

ON THE CITRIC ACID METABOLISM IN MAMMALS

prcparnlion (Fig. JO: HI, IV); the C/s certainly falls at first,
hm no doubt this only depends on a leveUing-out of the con-
centmlion over the whole preparation, after which it remains
stationary at high concentration only to rise again afterwards.

In such prejiarations the experimental conditions are not
ejuite constant for any considerable time, since blood pressure
.ind respiration arc influenced rather early, I have therefore
conducted .some perfusion experiments on isolated hindlegs of

c/s yA^

Fir/, 10. C/s after abdominal rvi.iccration.

I. Cn(, 1.7 kp. Pcnwclon 0.7 cc/kg. Conliniioiis injcclion of ndrennline,
(olnl O.i ing.

n. Rnbliil, .a kg. UrcUinnc 1 gm/kg. Continuous injcclion of glucose
0.4 grti per Jir.

HI. cm, .a.f! kg. Pernoclon 0.8 cc/kg. Adrcnnlinc O.l mg. Intrnv. injection
of US' mg Cilkg.

IV. Rabbit, 3.S kg. Pernoclon 1 cc/kg. Adrenaline O.l mg. Intrav. injection
of 27 mg Cilkg.

rabbits and cats, in whicli experiments the o.xygenation and
flow volume of the blood can be kept more constant and effects
from other organs (lung, heart, brain) avoided.

For perfusion of the hind-lcg.s I used the same apparatus as for liver
perfusion, but kept the circuit closed in order to ensure sufficient pressure.
I had access to a Dalc-Schuslcr pump, by which the flow volume per
minute, and with it the pressure, can be readily adjusted during the ex-

JOHAN MARTENSSON

perimeni without interruption of the circulation. The lateral arterial
branches were carefully ligatured, as were also all the vessels to the
bladder and genitalia, and the entire posteriors were constricted at the
beginning of the perfusion by means of a steel wire with a screw device
with which the vertebral column could also be compressed. Otherwise
the preparation leaks up into the anterior part of the body.

Of the experiments, the following may be submitted: 26. 2. 40. Rabbit,
4.35 kg. Urethane 1 gm. per kg. Perfusion fluid: 240 cc. heparinated blood
with about 20 % Tyrode’s sol.; haematocrit value 29 %. Min vol. 56 — 60 cc.
Pressure 60 — 65 mm. Hg. Pulsation 70 p. min. Temp. 37.6°. The C/s
of the perfusion fluid rose from 90.9 7 to 103 7 per cc. during a 50 minutes’
period. After addition of 45 mgm. Ci the C/s sank from 348' 7 immediately
after the addition to 308 7 per cc. after 20 min. and 282 7 after 50 min.

In a similar experiment on a rabbit the C/s rose in 80 minutes from

73.1 7 to 105 7 per cc., and in one on a cat it rose in 45 minutes from
41.7 7 to 55.3 7. After adding Ci to another hindleg preparation of a rabbit
the C/s was as follows:

5 15 30 50 min. after Ci addition

C/s 524 416 374 397 7/cc.

That the rises were actually produced by Ci, and not by, c. g., hexose-
diphosphoric acid, formation of which in the muscle was conceivable under
these conditions (Deoticke & Hollmann (1939), was verified in one of
the above-cited cases by the pentabromacetone method: the C/s rose from

67.2 7 to 84 7. Nor did the curves for the Ci estimation according to
Thunberg deviate from the usual type, and the “level” did not reach
lower than that of the standard curve.

The perfusion experiments, therefore, fully confirm the
results obtained from eviscerated preparations: the Ci-content
of the perfusion fluid increases during the experiment, and on
adding Ci up to a high concentration no undoubted oxidation
of Ci in the muscle is found.

Now this increase could depend upon the Ci-formation in
the muscle here being brought about, or being accelerated,
under the influence of other metabolic products tliat accumu-
late in a muscle preparation. Ci is also formed, however, in
muscle with normal metabolism. In one series of experiments
I took samples at the same time from the arterial blood and
the peripheral part of the femoral vein of an otherwise intact
animal. The results are furnished in the table given below and
show that the C/s of the blood flowing out of the muscle is
higher than that of the inflowing blood.

The animals were narcotized as usual. The venous blood was drawn
by puncturing the femoral vein immediately after the latter had been

ON THE CITRIC . ACID METABOLISM IN MAMMALS

centrally clamped. The arterial blood was taken from the femoral artery
on the other side, or from the carotid. (The difference is slated in
percentages of C/s of the arterial blood.)

Animal

Art.

V.fem.

Diff.

Animal

Art.

V. fern.

Diff. =/

Cat

40.7

33.0

30.2

Rabbit

109.0

113.0

3.6

))

38.2

46.3

21.2

71.4

77.7

8.8

57.8

66.2

li.5

42.7

45.0

5-4

55.0

63.9

16.2

42.5

43.5

2.4

Dog

53.2

58.3

9.0

66.0

70.4

5.7

Rabbit

77.3

80.8

4.6

62.0

64.1

3.4

In rabbits the difference is distinctly smaller than in cals, but a higher
Ci concentration is nevertheless always found in the venous blood. Sub-
stantial quantities of Ci, however, can be brought along with the copious
blood flow from the total musculature of the body and 5 *et the difference
in Ci-conlcnt between arterial and venous blood may be small.

In heart muscle the behaviour of the Ci metabolism, does
not appear to be materially different from what it is in ordinary
muscle. A perfusion test on a rabbit’s heart failed to reveal
any change in the Ci-content of the perfusion blood during the
experiment.

The heart, which weighed 9.4 grams, was perfused from the aorta, m
the usual Langendorff preparation, with 160 r.c. of undiluted heparmated
blood. The pressure was kept at 105 mm. Hg, which gave a blood flow
of 29—26 cc. per minute; the heart beat well the whole time. For a peno
of 40 minutes the G/s remained unchanged, initial value 39.1 V per cc.,
final 39.2 After addition of Ci the C/s was at first 282 y per cc., a er
40 minutes 274 '/. The circulating quantity of Wood was certainly large
proportionately to the weight of the heart, for which reason small changes
did not reveal themselves, but the blood can none the less be assumed to
have passed through the heart muscle six or seven times during eac

period of the experiment.

The Oitric-Acid Excretion in the urine.

The experiments hitherto described point to the kidneys as
being the organ by which Ci is eliminated from the body.

According to the literature reviewed earlier it
two per cent, of the amount of Ci supplied by mouth is
urine After parenteral administration the urinary Ci excretion ma> be

JOHAN MARTENSSON

considerably larger: Salant & Wise (1916) had found 30 % in experiments
on the cat, Orten & Smith (1937) up to 40 % in the dog after such large
intravenous doses as 400 mgm, of sodium citrate per kg. b. w. The ex-
cretion is accordingly bound up, as might be expected, with the actual
Ci concentration in the blood: This is very vigorously augmented after

Fig. 11. CIs and urinarg Ci excretion after intravenous injection of Ci.

I. Cat, 2.4 kg. Urethane 1.23 gm/kg. Intrav. injection of 100 mg CUkg.
Ci-conccntration in urine (white field) : before Ci-injection 69 y, in first
portion 18450 y, in second 5890 y, and in third 1480 y/cc. Total Ci-
excretion 103 mg or 43 % of the amount administered.

II. Rabbit, 3.9 kg. Urethane 1.25 gra/kg. Intrav. injection of 100 mg Cilkg.
Ci-conccntration in urine (black field): before inj. 188 y, in first portion
11520 y, in second 2520 y, in third 1530 y/cc. Total Ci-excretion 19 mg
or 4.9 % of the amount administered.

intravenous injection of Ci, whereas after oral administration it does not
attain such a height that the renal threshold is materially exceeded.

The most important precaution in a tolerance test designed
to investigate the Ci excretion is to follow the C/s and to collect
the urine at sufficiently short intcrv'als, as the Ci elimination

ON THE CITRIC ACID METABOLISM IN MAMMALS 55

is SO transitory. Two typical experiments of this kind on rabbit
and cat are described in Fig. il and its accompanying text.
These show that the kidneys possess an exceedingly high power
to concentrate Ci in the urine. The most extreme case was in
a dog, in which continuous infusion of Ci gave a urinary con-
centration ot not less than 3.8 %, This excessive rise of the Ci
concentration in the urine presupposes a powerful elevation of
the C/Sf and it vanishes as rapidly as the Cjs sinks. Hence even
a powerful increase in the concentration need not at all become
v isible if the Ci estimation is made on, say, samples of daily
amounts; Salant & Wise found no Ci in the urine after in-
travenous injection of 100 mgm per kg., just the dose used in
the experiments recorded here. The experiments confirm the
old observation that in cats the Ci excretion is considerably
larger than in rabbits, 43 % and 4.9 % respectively of the same
(!i dose. In two other cases the excretion in the cat was 27 %
and 30 in four cases among rabbits 3.8 %, 8.2 %, 4.i % and
4.8 %. This divergence must be due to a primary difference
in the kidneys themselves, since the C/s keeps high for a longer
time in the rabbit and might accordingly be expected to cause
a larger excretion.

In the experiments ttie urine was collected through a cannula tied into
the top of the bladder. By this simple means urine could be continuously
collected even from rabbits, on which animals several earlier experiments
had failed. Manipulations of the urinary organs are apt to give rise to
oliguria or complete anuria through ureteral spasm or other causes, and
this doubtless reacts on the passage of blood in the kidneys, since injected
Ci is then not eliminated from the blood in a normal manner.

In an experiment on man one gram of sodium citrate was intravenously
injected during four minutes, and the C/s rose from the normal 25 7
per cc. to 116 7 per cc. The collected urine from the first 45 minutes
showed a Ci concentration of IGOO 7 per cc., a moderate rise.

The Citric Acid Metabolism in Perfused Kidney.

If the kidneys are to answer for the greater part of the body’s
removal of administered Ci, it must be assumed that, collateral
with the urinary Ci excretion, a considerable oxidation of Ci
goes on in these organs. To investigate this aspect more closely
I made a series of perfusions on isolated cat kidneys with hepa-
rinated blood. With this object the perfusion can be under-
taken direct, without having a lung preparation, which other-

JOHAN MArTBNSSON

Avise senses to “detoxicate” the blood of vasoconstricting sub:
stances.

The kidneys were left in situ. The aorta was clamped liclow the renal
arteries and the inflow cannula was attached there. The outflow cannula
was inserted in the V. cava inf. below the renal veins, whereupon the
former Avas ligatured above the latter, and the cat was allowed to bleed
through the renal vessels. Perfusion was then started, and simultaneously
the. aorta was ligated above the renal arteries, by Avhich means the

c/s y/cc

M?n.

Fig. 12. The CIs during kidney perfusion.

I. Cat, 2.C kg. Pernocton 0.7 cc/kg. Perfusion fluid 260 cc heparinated
blood (with about 35 % Tyrode’s sol.). Addition of 86 mg Ci. Lively
urine secretion during the Avholc experiment, about 17 mg Ci, or 20 %
of the amount supplied, being excreted.

II. Cat, 4.4 kg. Pernocton 0.7 cc/kg. Perfusion fluid 2G0 cc heparinated
blood (with about 25 % Tyrode’s sol.). Addition of 107 mg Ci. Urine
secretion only at the outset, 5.3 mg Ci being then excreted.

III. Cat, 3.6 kg. Urethane 1.6 gm/kg. Perfusion fluid 300 cc heparinated
blood (with about 35 % Tyrode’s sol.). Addition of 107 mg Ci. Only
insignificant urine secretion to begin with.

circulation thtough the kidneys could be left, practically speaking, un-
interrupted. All lateral branches and the vessels to the suprarenals and
left spermatic vein were carefully lied. The urine was collected through
a cannula tied into the top of the bladder.

Whether or not secretion of urine will be able to proceed appears to

ox THE CITRIC ACID METABOLISM, IN MAMMALS 57

cpend lo a high degree on whether volume of flow per minute and
pressure arc correctly adjusted at the outset of the experiment. This
secretion obviously plays a minor part in e.Tperiments on the Ci-oxidation
in the renal parenchyma, but when the effect of different factors on the
Cl excretion has to he studied the flow rate and pressure must be stabilized
at the .start and then kept constant.

Fig, 12 gives quaniilative data from three experiments with a
single bulk addition of Ci. Even if urine secretion has ceased, so
that no Ci can disappear by that path, yet Ci is rapidly eliini-
nated from the blood. This shows that the Ci-oxidation in the
renal parcnchijma is very active. The conversion continues
at a relatively high rate even down in the low concentrations,
and one is therefore entitled to assume that it is the kidneys
which keep the Cjs at a constantly low level, and that the oxi-
dation of Ci there is alone sufficient to account for the body’s
power to deal with the Ci supplied to it.

The amount of Ci oxidized in the renal parenchyma is most
accurately calculated on the basis of experiments with contin-
uous administration of Ci, as the C/s is not then raised exces-
sively. A comparison of three experiments, which were carried
out under the same conditions to ensure fully comparable
results, gave a value of 32 y per minute per gram kidney when
the C/.S averaged 90 y per cc.

The cal kidneys were perfused with heparinated blood with 20 — 25 %
addition of T 3 ’rodc’s solution. After a preliminary period, when the blood
flow per minute and pressure had been stabilized, a sample of the per-
fusion fluid was taken, whereupon sodium-citrate solution equivalent to
40 nigm. of Ci was added in uniform drip during 20 minutes. A fresh
sample was then taken; the C/s at beginning and end of the Ci-drip is
stated under the heading “Ci-conc.” in the following report of the ex-
periments.

■

Min. Vol.

Pressure
mm Hg

Ci-coiic.

7 /cc

Wt. of
Kidney

Ci Consump-
tion in mgm

25.7—160

18.1

29.7

36.5—123

17.2

31.8

III

63.5-136

29.7

34.5

In order to verify that the kidneys in situ play the same part
in the normal Ci-metabolism, I drew a series of samples from
the arterial blood and the renal vein at the same time. In most
cases the blood was taken direct by puncture of the renal vein,

JOHAN mArTENSSON

in a couple of cases from the left ovarian vein, which during
pregnane}' is well developed. The result is given in the follow-
ing table, the differences being calculated in percentages of
the arterial blood values.

Animal

Arterial

Renal V.

DilT. %

Animal

Arterial

Renal V.

Diir. X

Rabbit

111

78.2

29.6

Cat

51.2

42.3

17.4

89.7

65.0

26.5

64.0

56.0

13.7

127

78.«

38.3

69.0

54.0

21.7

103

71.0

31.1

55.0

42.1

23.5

))

83.0

66.3

26.5

56.4

44.6

20.0

108

73,1

30.5

Dog

53.2

37.0

30.5

80.8

55.0

31.0

80.7

46.4

42,5

68.0

44.8

35.0

57.6

25.5

55,7 ^

))

68.3

43.7

36.0

Rabbit

46.5

36.0

24.3 *

^ By the pcniabromacetone method.

The tests show that there is a marked difference between
the Ci-concentration in arterial and renal-vein blood even under
normal conditions, when the C/s has not been raised by admi-
nistration of Gi and the Ci-excretion is consequently very small.
The difference is greater in the rabbit, about 30 % , than in the
cat, about 20 %. But this is counterbalanced by the relatively
much smaller weight of the kidney in the rabbit, on which
account a lower renal blood flow per unit time has also to be
taken into consideration. As a matter of fact, the Ci-consump-
tion per minute per gram kidney of the rabbit may be regarded
as being of the same order as that of the cat. This is shown
by the experiments described below, in which the Ci-oxidation
was evaluated on the basis of the Ci-differences found in arter-
ial and renal-vein blood and the volume of blood passing
through the kidneys per minute.

Attempfs to perfuse isolated rabbit kidneys ■with the same technique ns
for the cat failed on account of vasoconstriction. Tlie Ci-consumplion
■R'as therefore measured in rabbit kidneys in the following manner: Into
the right jugular vein of a liepnrinated rabbit a coarse, bent cannula %vns
inserted and connected by a rubber tube with a cannula that fitted into
the left renal vein. The connexion was filled with Ringer’s solution. The
renal vein was ligatured, the cannula placed quickly in, whereupon
circulation could immediately commence at the anastomosis. A T-lube

ON THE CITRIC ACID METABOLISM IN MAMMALS

was filled in llie taller for sampling. Immedialcly after, a sample was
taken from the carotid arlerj’. The flow volume per minute was measured
by allowing the blood lo flow out into a measuring-glass and noting the
number of seconds required for 5 cc. to collect; the blood was afterwards
rc-injcctcd. This technique is successful with rabbits, but in cals a
transudation is likclj* to occur under the renal capsule. As the ex-
periments were not originally intended for calculation of the absolute
Ci-consumplion, no hacmalocril reading was made. In the following
approximate calculation the Gi-contenl of the blood waS considered to be
70 9S of the C/s; the Ci-cxcrction in the urine is so small that it can be
neglected.

A. c.arot.
C/s y/cc.

V. renal
C/s v/cc.

Min. Vol.
cc.

Ci-consump-
tion p. min.
p. gm.

Rabbit

91.0

G9.1

21.4

107

82.6

43.2

9fi.<

70.2

3G.t

151

7.8

34.5

The Citric Acid Content of Serum after Nephrectomy.

Previous experiments would lead one lo expect a rise in the
C/s if Ci-elimination via the kidneys were prevented. Such
proved to he the case, as can be seen from the curves in Fig. 13.

Tlie animals were under narcosis all the time. Nephrectomy was per-
formed by way of laparotomy' or from the back. The blood samples from
rats were taken from different animals at varying lengths of time after
ligation of the renal vessels, since continuous samples of sufficient size
cannot be drawn from the same animal.

The curves show that in rabbits and rats the C/s rises veiy
rapidly up to four or five times the normal value, while in
cats the rise is much less, only double the normal figure. 1 he
cause of this may be that after the C/s has reached a certain
elevation the liver of the cat, having greater Ci-oxidizing power,
is able to maintain an equilibrium between Ci-formation and
Ci-breakdown. There is also the possibility that an acidosis

sets in and counteracts the rise.

The narcotized rabbits died only three or four hours after
the nephrectomy, in spite of the fact that only a few bloo
samples bad been drawn. The narcosis must in some way
be responsible for this, for, if the kidneys are extirpated under
ether narcosis and the rabbits afterwards allowed to revive,

JOHAN MARTENSSON

they survive five to seven days. Then, however, first conies
the postoperative or postnarcotiC fall of the C/s observed both
in man (Schersten, 1931) and rabbits (GrOnvall, 1937), and
after that the rise takes place much slower than in narcotized
animals: the same rise will take days instead of hours (Fig. H).

c/s y / c & C/s

Fig. 13. The Cls after nephrecfomtj, narcotized animals.

I. Cat, 3.G kg. Urethane 1.25 gmfkg.

II. Rabbit, 2.7 kg. Pernocton 0.95 cc/kg,

III. Rats, 0.2 kg. Ether narcosis.

Fig, 14. The Cls after nephrectomy, nan-narcotized animals.

I. Rabbit, 4.2 kg. II. Rabbit, 6.2 kg.

The disturbance in metabolism brought about by nephrectomy is
evidently very strongly accentuated by the derangement narcosis involves.
Narcosis is stated to give rise to a certain amount of acidosis and an
elevation of blood urea; furthermore, after nephrectomy blood urea and
other non-protein compounds have been observed to increase more rapidly
in dogs under narcosis (quoted from BEECHEn, 1938). Anoo.v and

ON THE CITRIC ACID METABOLISM IN MAMMALS 61

Gisselsson (1938) observed the same difference between narcotized and
non-narcotized animals in respect of creatinaenpa, phosphataemia, and
acidosis after nephrectomy.

The Citric Acid Concentration in Various Tissues Norm-
ally and After Administration of Citric Acid.

My experiments have shown that the Ci-oxidation is effected
mainly in the kidneys, w'hile the liver has little and the muscle
no power to break down Ci. This is in conflict with the results
of numerous works on Ci-metabolism in minced tissues (inter
alia Langecker 1934, Breusch, 1937), according to which the
Ci-metabolism is greatest in liver, less in kidney and muscle.
Later, Krebs and Eggleston (1938) have shown that the
breakdown of Ci in muscle, where the iso-citricodehydrogenase
is very labile, is under optimum conditions still greater than
in the liver, which furtlier accentuates the lack of agreement
between in-vivo and in-vitro experiments. However, it must
be borne in mind that the dehydrogenase reactions are revers-
ible and that in the intact cell they may be differently linked
than in the injured cell. Another possibility is that the cells of
the intact organ have plenty of other fuel to attack before
beginning on Ci.

In connexion wiUi these discrepancies between experiments
on intact and minced tissues, the most important question is,
however, whether the substance in question is able to penetrate
into the uninjured cells. This, I have endeavoured to discover,
so far as Ci is concerned, by investigating the Ci-content of
the tissues before and after administration of Ci to the animals.

For the Ci-estimations I used the pentabromacetone method according
to PucHER, Sherman and Vickery (1936), and followed their directions
for the treatment of the tissues. These authors give the following Ci
concentrations in tissues from dogs: kidney 12 7 per gm., muscle 11 7,
and liver only 1 7. However, I soon found that the Ci-value obtained when
the tissues were immediately immersed in trichloracetic acid and pounded
very rapidly was materially different from that obtained when they were
first cut up with scissors or treated less rapidly. This applies especially
to liver and kidney; for muscle it plays a minor part. In a piece of liver
that had been finely cut during one minute, the Ci-concentration fell from
19 7 per gm. to 8 . 6 . This may explain why Langecker (1934) could not
find Ci in the tissues and why, as a general rule, such low concentrations
have been found in the liver — the finding that is accepted generaUy as
evidence that the chief seat of the body’s Ci-oxidation is the liver. Orten

JOHAN mArTENSSON

& Smith (1937) froze down the tissues for an investigation on rats, which
procedure ought to be effective, but as the animals were first bled to death,
these authors may nevertheless have obtained too low values. If the
animals are killed and the circulation through the organs is thus cut off
before samples are taken, I found that the concentration in liver and
kidney falls considerably in even a couple of minutes. (In one such
experiment the amount of Ci was reduced from 26.5 y to 12.7 y per gm.
in the liver, from 77 y to 25.7 y per gm. in the kidney.) I therefore took
the specimens of the organs with circulation undisturbed or else took them
practically simultaneously with the ligation of the vessels. They were
dried and squeezed as free as possible from blood in a compress before
being immersed in the trichloracetic acid and ground with sand. The
kidneys were first clipped in two, so that the capsule could be quickly
removed and the pelvis dried free from urine. When two series of samples
were taken from the same anithal, a whole liver-lobe, a kidney and the
thigh-muscle were taken from one side in the first round, and tissues from
the other side, which were llius still quite intact, in the second round.

Some illustrative results from these estimations of the Ci-
concenlration in liver, kidney and muscle before and after
administration of Ci have been collocated in the following table.
The value for blood serum is also given as a measure of the
Ci-concentration in the medium with which the cells are washed
and have to enter into equilibrium. (The Ci concentration is
given in y per gram).

Blood

Serum

Liver

Kidney

Muscle

Ci-nddition

Rabbit

46.3

19.0

88.0

19.5

None

))

6.S.5

26,3

77.0

22,.3

III

325

48.0

570

57.7

100 mg/kg

940

131

945

64.0

1.32 »

720

G.5.0

680

50.8

(20 min. later)

Cat

.54.8 ')

25.3

55.8

29.2

None

640 »)

93.8

2300

73.1

120 mg/kg <27 min.)

800

136

1860

60.1

100 mg/kg

464

54.1

1300

58.8

(20 min, later) j

^ Estimated by the Thu.nberg method.

It will be seen that the Ci concentration in liver and muscle
is normally lower than in blood serum, while in kidney it i.s
the same or higher. Following addition of Ci the Ci-contenf
of liver and muscle rises only slightly in relation to the C/s.

ON THE CITRIC ACID METABOLISM IN MAMMALS 63

(It may be regarded as doubtful whether the increase found
is much higher than is accounted for by the Ci-content of the
blood left in the tissues). This also applies if the C/s has been
elevated for a relatively long time, such as in three of the
above-recorded experiments. On the other hand, the concentra-
tion in the organs declines with falling C/s, although this is
still very high. The position is otherwise with the kidneys.
In rabbits the Ci-concentration increases at least as much as
in serum, and in cats a very powerful accumulation of Ci taltes
place in the renal tissue.

As to how the re-absorption conditions for a substance oxi-
dized in the renal parenchyma itself are to be conceived in
detail, I do not venture here to take up a position. It is how-
ever tempting to suppose that the ability of the renal tissue to
o.xidize Ci becomes a limiting factor for the re-absorption. This
ability is obviously augmented with rising Ci-concentration, but
when it is exceeded, Ci accumulates in the renal parenchyma
in such quantities that re-absorption is hindered and large
amounts pass out with the urine.

Since the Ci-concentration in both urine and renal paren-
chyma is considerably lower in the rabbit than in the cat, it
is conceivable that Ci filters out at a lower rate in the rabbit,
even if the C/s is the same.

Manifestly the high Ci-concentration in the renal tissue may to some
extent depend on Ci-rich urine being left in the uriniferous tubules, in
spite of the pelvis having been cut away and the tissue squeezed out, but
this residue can scarcely be so large that it can even approximately ex-
plain the Ci-increase.

Heart Muscle. This differs very materially from ordinary
muscle, in that the Ci-concentration is normally considerably
higher than in the serum. After Ci-administration, however,
the Ci of heart muscle does not increase in any high degree.

Normal |
C/s 7/cc

Heart Muscle
7/gra

Rabbit j

42.7 >)

125

66.0

265

61.5

233 1

Cat

43.5

130

C/s after
Ci-addition

Heart Muscle
r/gm

Rabbit

325

250

720

320

Cat

640

206

464

203

‘ Estimated by the Thunberg method.

JOHAN MARTENSSON

In one case the Ci of a working rabbit-muscle was determined. A
group of dorsal thigh muscles, which had to work against an elastic
resistance, were e.xcited from the nerve to make 45 jerks per minute
during 30 minutes. Tlie Ci-concentration was 39 y per gm., in the corres-
ponding group at rest on the other thigh it was 41.8 y per gm., thus no
sure difference.

From the experiments the conclusion may he drawn that
Ci has little or no power of penetrating into intact hepatic or
muscular tissue. Certainly the low concentration in the organs
after addition of Ci might be due to a rapid oxidation of any Ci
that had intruded (as is proved to be the case in the liver),
but, since liver or muscle tissue removes no Ci or very little
from the blood in perfusion experiments, in which Ci is offered
to the cells in so physiological a way as via the capillaries, the
conclusion put forward above may be regarded as well-founded.

Ci penetrates very readily into the renal parenchyma, but it
must be remembered that its path there is via a re-absorption
through the tubular epithelium, since the concentration in the
parenchyma may be so much higher than in the blood. Still,
judging from perfusion tests, it should be remarked that the
oxidation of Ci in the renal tissue appears to be equally active
even when secretion of urine ceases.

Tlie rapid fall in the C/s after intravenous injection of Ci
(Fig. 1) seems to receive an adequate explanation from these
experiments: In addition to the large Ci-excretion in the urine
when the C/s is high, there is this large accumulation of Ci in
the renal parenchyma with its resultant higher rate of oxida-
tion.

These experiments to determine the ability of Ci to penetrate
into the tissues would also appear to have provided a satisfac-
tory explanation of the discrepancy between in-vivo and in-vitro
experiments. For I have not found any indications that the
Ci formed in the body behaves differently from that supplied
from without. That in respect of diffusion the endogenous Ci
does not differ in behaviour from the exogenous is evidenced
by the following experiment, in which an elevation of the C/s
was brought about by means of bilateral nephrectomy.

22. 3. 40. Rabbit, 4 kg. starved 48 hour.?. Urethane 1.25 gm. per kg.
Bilateral nephrectomy. After 2 Vs hours samples were taken from liver,
muscle, and arterial blood and were analysed by the pentabromacctone
method. The C/s had risen to 200 y per cc., but the concentration in the
liver was only 78.5 y per cc. and in the muscle 52 y.

ON THE CITRIC ACID METABOLISM IN MAMMALS 65

Lynen and. Neciullah (1939) have studied the breakdown of succinic
acid, malic acid and Ci bj’ yeasts. They found that a number of earlier
disagreements maj' be reconciled by faking info consideration the fact that
the wall of the ycasf-ccll is but little permeable to these acids. On the
other hand, according to Harrison (1939), succinic acid penetrates readily
info a frog s sarforiiis, although it does not undergo consumption there.
And Lundsgaard (1938) found that alcohol is not consumed on muscle
perfusion, although it makes its way info the cells and these contain the
specific dehydrogenase.

Discussion of the Experimental Results ’.

The Ci in the circulating blood has a definite tendency to
remain at a constant level. There accordingly exists a constant
equilibrium between the Gi-intake and Ci-ontput of the blood.

^ After this work had been written, a preliminarj' report was published
bj' Dickens (1910), who seems to have opened up new fields of inquiry
respecting the significance of Ci in the body under normal and pathological
conditions. While investigating the Ci-concentralion in different tissues
Dickens found that most of the body’s Ci is contained in the skeleton, in
which it may amount to not less than 1000 mgro. per 100 gm. of the dry,
de-fatfed substance. TIius there is here a large store of Ci, presumably as
Ca salt. This discovery' adds interest to the works dealing with the relation
between Gi-metabolism and calcium metabolism, especially the conducive
influence Ci has on the retention of calcium (e. g. WesterlunT) 1931,
Lanford, 1939) and with the beneficial effect Ci has on experimental
rbachilis (c. g. Hathaway and Meyer, 1939). It also provides us with
some new points of view respecting the use of Ci as an addition to cow
milk in artificial feeding and, in general, as a beneficial agent in early
infancy (see Siwe, 1938). It remains to be seen whether the Ci in the
skeleton takes part in the constant Ci-balance in the circulating blood. The
conditions under which my' experiments on eviscerated preparations and
my hindlegs perfusions were carried out do not exclude the possibility
that the Ci which I thought came from muscle actually came from the
skeleton instead. As to the rest of my experiments, a possible flow of Ci
from the skeleton cannot play any part. Dickens further found that
malignant tumours from mice, rats and rabbits contain much more Ci
than normal (issue and that embryonic tissue does so too. Considered
along with ScHERSTfiN’s finding that the male gonads are very rich in Ci,
this may suggest that a high Ci concentration occurs where a brisk new-
growth of tissue elements is going on. The fact that (he C/s is higher
in foetuses and young animals has already been demonstrated y

GrOnvael (1937). . .

In tests with the enzymic method Thunberg (personal communication)
has not only confirmed Dickens’ find respecting Ci in bone, but found
still higher values (up to 2 %).

JOHAN MARTENSSON

According to the experiments reported here, the Ci is mainly
supplied from the portal area and from the muscles, while it
is eliminated to a large extent by the kidneys and to a small
extent by the liver, more in cats than in rabbits. Tolerance
tests show that the bodj’^’s power to transform Gi supplied from
without is much greater than is normally called on. My experi-
ments have been largely carried out on isolated organs, but I
have consistently tried to verify that these organs when in situ
play the same part in the Ci-metabolism. When Ci-oxidation
has been referred to here, this has not meant, of course, the
complete combustion of Ci to carbon dioxide and water, but
only its breakdown into some other compound. I have not
registered the influence of Ci on the Os-consumption or the
respiratoiy quotient of the organs.

The object of the following account is to put my experimental
results into relation with that conception of the formation,
breakdown and significance of Ci in the body which has
appeared in the literature of recent years.

Knoop and Mautius (193G) consider that Ci can he formed from oxalo-
acetic acid and pyruvic acid, and- they hold it probable that the synthesis
takes place in vivo in the same way. This has been corroborated by
Breuscii (1939). The Ci-breakdown is assumed to occur in the following
manner (Martius, 1937, 1938): Ci is first transferred to the unsaturated
cis-aconitic acid, which takes up water again and forms isocilric acid.
This is dehydrogenated to oxalasuccinic acid, which spontaneously gives
off CO 2 and forms a-kctoglularic acid, which under decarboxylation and
dehydrogenation is converted into succinic acid and later via fumaric acid-
malic acid into oxaloacetic acid.

I have not gone here into the question of from what Ci is
formed and what the more immediate products of its break-
down are in the body. To give reliable results an investigation
into this aspect would demand parallel determinations of sev-
eral of the above-mentioned metabolites, for which there are
only tedious analytical methods that are not sufficiently ac-
curate for the concentrations normally occurring in the body.
Moreover, there would first be required a similar investigation
into the normal metabolism of these acids in the body as that
here accorded Ci, this as a safeguard against false interpreta-
tions.

The chain of Ci-breakdown products would thus include nil the acids
wliich were shown by Tiiu.vberg 30 years ago to have a distinctive position
in the metabolic processes of minced muscle. Tliey liave since generally
been regarded as disintegration stages \yhich, in the course of oxidation.

ON THE CITRIC ACID METABOLISM IN MAMMALS 67

are conlinually being formed and continually being broken down again.
Through the work of Szent-CyOrgyi and collaborators (1935), however,
Ci-dicarbo.xylic acids arc assigned another metabolic role- These acids are
thought to act as catalysts in the oxidation of carbohydrates, forming an
intermediary link for the transport of hydrogen to the cytochrome system.
In this they are conceived as reversiblj’ swinging between the hydrogenated
and dehydrogenated form, without entering themselves into the chain of
reactions. SZENT-CYOnGYi’s theories, however, involve certain difficulties
of interpretation, as was pointed out c. g. by Martius (1939). Later, Krebs
and Joif.vsoN (1937) set up another theory, based on the new conception
of the formation and breakdown of Ci. According to this — the “citric
acid cycle’’ theory of Kreiis — , Ci is formed in the muscles from oxalo-
acetic acid and a product of carbohydrate metabolism, a triose or similar
combination. When Ci is broken down later, oxaloacetic acid, according
to the course of reaction outlined above, is rebuilt and with triose forms
Ci again. In this way a cyclic process is maintained, the result of which
is that the carbohydrate participating in the reaction undergoes complete
combustion. Accordingly, it is here thought that the reaction proceeds
in only one direction and that Ci as well as the Cj-dicarbo.x}’lic acids take
part in the chain; thus the theory differs in principle from Szent-GyOrgyi’s.
On these lines one would get a plausible explanation of the extremely
high activity of the citricodchydrogenase and of its universal occurrence
in the cells. The low but rather constant Ci-concentration in blood and
tissues — in man scarcely more than O.oooi mol. — would also seem to
argue in favour of a catalytic function. The Gi-cycle theory is also further
supported in recent works of Krebs (1937), Krebs, Salvin and Johnson
(1938), and Krebs and Eggleston (1938). The last-mentioned authors
think that insulin has its point of attack in this Ci-cycle. The theory has
however met with opposition from several investigators, e. g. Breusch
(1937, 1939), Banga, Ochoa and Peters (1939), Thomas (1939). Stadie,
Zapp and LukenS (1940) were not able to confirm the catalytic function
of Ci or the insulin effect on the Ci-cyclc. No conclusive evidence for
either of these theories seems to have been advanced as yet.

On the basis of my experiments no attitude can be directly
taken up towards these theories based on in-vitro experiments.
Any Ci metabolism that might be going on within the cells
would not necessarily find expression in my experiments. Even
if Ci cannot pass tlirough the cell- wall, or can only do so wnth
difficulty, a Ci-formation is conceivable from substances that
can pass through. On the other hand, the cell-wall ®

permeable to cleavage products of Ci but the Ci-meta o isna
itself may go on within the cells and need not become app^en
in perfusion tests. In this way the cells can nevertheless have
use for their highly active enzyme.

JOHAN MARTENSSON

This question, as a matter of fact, is v^ery difficult to solve: by
perfusion experiments one cannot expect with any certainty to
get at intracellular processes, by in-vitro experiments on minced
tissues one can obtain a good idea of the metabolic possibilities
possessed the cells but no reliable knowledge of how the
cells make use of these possibilities, as they are seated linked
together in an intact complex. Moreover, a considerable part
of the enzymic potentiality of the cells is doubtless only called
upon in pathologic situations, e. g. autolysis.

However, I have been able to take up a position with regard
to a series of in-vivo experiments which have been taken as
evidence confirming the citric-acid-cycle theory of Krebs.
Orten and Smith (1937 a, b) were able to show that the Ci-
excretion in the urine of dogs and rats increases immensely
after intravenous injection in large amount of certain metabo-
lites, chiefly malic acid, fumaric acid, and succinic acid. But
such substances as malonic acid and maleic acid also raise the
Ci-excretion in the same degree. Later experiments (Orten
and Snhth, 1938 a, b, 1939) led these authors more and more
to the opinion that the Ci is formed from the injected substances
and that this formation takes place chiefly in the kidney or is
dependent upon the presence of the kidney. The kidney was
thus assigned a role in the Ci-metabolism that is the direct
opposite of the one indicated by my experiments. Orten and
Smith’s experiments were fully confirmed by Krebs, Salvin
and Johnson (1938) and were accepted by these authors as
proof that it is possible to demonstrate the occurrence of the
citric acid cycle in the li\ing organism. However, by de-
termining the Ci-concentralion in samples taken at the same
time from the arterial and renal-vein blood, and b}’’ means
of kidney perfusion tests, I was able to show that Ci is not
formed in the kidneys after administration of malic acid, but
that this acid inhibits the normal breakdown of Ci in the renal
parenchyma and thereby brings about a powerful increase in
the Ci-excretion in the urine (MArtensson, 1939). It is this
increase in the excretion of Ci, caused by the inhibition of the
Ci-breakdown in the kidneys, that has been interpreted as a
formation of Ci from the injected substances. Consequently,
the in-vivo experiments cited above cannot be accepted as proof
of the accuracy of the Ci-cycle theory of Krebs.

The effect of the injected substances on the Ci-excretion is
easier to understand in the light of my later experiments, which

ON THE CITRIC ACID METABOLISM IN MAMMALS 69

show that the Ci-concentration in the renal tissue is high and
increases markedly on administration of Ci. It is very possible
that the other metabolites accumulate in the renal tissue in the
same way as Ci. In the works cited they were supplied in
very large doses, and it is therefore easy to understand that
they ought to have a strong effect, whether this effect is inter-
preted as operating through their specific adsorption to, and
blockage of, the isocitricodehydrogenase (e. g. the effect of
malonic acid on succiriodehydrogenase, Quastel and Woold-
ridge, 1928) or through the dehydrogenating capacity of the
renal parenchyma being quite insufficient to cope with the Ci
when this tissue is charged witli some other readily com-
bustible substance to a high concentration. Favouring the
first alternative is the fact that a pure “enzyme poison” such as
malonic acid has this effect; it also inhibits the oxidation of
Ci in the tissues (Breusch, 1937). Another support for this
alternative is the somewhat similar constitution of malic acid
and isocitric acid, as was pointed out to me by Lehmann, who
(1938) investigated similar problems relating to the lactico-
dehydrogenase.

In experiments on the dehydrogenation of Ci by cucumber-seed extract
I have found a retardation of the MB-decolorization to follow addition of
1-malic acid in concentrations not exceeding those in which malonic acid
causes an inhibition. Tliis is mentioned as a parallel. In order to prove
anything the experiments would have to be carried out with isocitrico-
dehydrogenase from kidney.

Since my first Ci-work appeared (MArtensson, 1938), in
which the Ci-concentration was shown to be lower in the renal-
vein blood than in the arterial blood. Smith, Orten, Johnston
and Banguess (1939) have conducted similar experiments on
dogs. Conformably with their conception of the course of
events, they had expected to find, under the influence of malic
acid, a higher concentration of Ci in the renal- vein blood than
in the arterial blood, due to a “reverse leakage” similar to what
occurs in the case of ammonium. They found the converse, a
lower concentration in the renal-vein blood, both before and
after malate injection. This is accordingly a good con irma
tion of my results, the more valuable because the experiments
were carried out on non-narcotized animals with the kidney
brought subcutaneously forward. On the other hand, m a
work of Smith and Meyer (1939) on the influence of diet
on the endogenous production of citric acid, alt ese me a

JOHAN MARTENSSON

bolites, whose influence on the Ci-excretion has been studied,
are spoken of as “definitely recognized precursors of citric
acid”, and Simola and Kosunen (1938) are stated to have
extended this series of precursors to pyruvic acid, a-ketoglutaric
acid, and glutaric acid. SiMOLA (1938) nevertheless empha-
sizes that the effect produced by these substances may depend
upon their preventing further cleavage of Ci, which therefore
accumulates.

My investigations would appear to have shown that all these
works on ‘‘the precursors of the endogenous citric acid” are in
no way conclusive, since their results may be entirely due to a
displacement in proportion between the Ci broken down in the
renal parenchyma and that excreted in the urine. In fact, it
is for the present an unsolved question whether certain of these
substances in experiments in vivo give rise to a Ci-formation
elsewhere than in the kidneys. That possibility manifestly
remains open, and for oxaloacetic acid and pyruvic acid is
fairly probable judging from in-vitro experiments (Martius,
1937; Breusch, 1939). Similar^, it must be said that all works
on the influence of diet on the endogenous Ci are not very
convincing so far as they postulate the derivation of Ci from
a particular kind of food (e. g. FCrth, Minnibeck, and Edel,
1934; Sherman, Mendel and Smith, 1936; Verkade, 1938,
quot. from Smith and Meyer, 1939). The fact is that the diet,
without giving rise to Ci, maj’^ have an influence on the quantity
of Ci in the urine (1) by affecting a bacterial Ci-production in
the intestine, (2) by direct action on the Ci-oxidation in the
renal parenchyma, and (3) by indirect action in the form of a
displacement of the acid-base equilibrium. These complicated
factors may, on one hand, entirely conceal, on the other, power-
fully accentuate, any Ci-formation going on from the food.

Breusch (1939) found that some Ci was formed in minced kidney after
addition of oxaloacetic' acid. He interprets this ns a mechanism by which
the body gets rid of surplus Ci-dicarho.xylic acids, which arc nett them-
selves excrctable in the urine, and by wliich it keeps the concentration of
these important acids constant in the tissues. Sucli an explanation would
also cast more light on the import of Orten and Smith’s results. This
broad theory of in-vivo conditions founded on an in-vitro experiment loses
however much in force in view of: (1) the fact that in the same work it is
shown that fumaric acid-malic acid cannot give rise to Ci, (2) the results
submitted here from my experiments, (3) the fact that at least the succinic
acid is excreted in the urine just as readily as citric acid (Fors.s.man, un-
published).

ON THE CITRIC ACID METABOLISM IN MAMMALS

Among the breakdown products of Ci, a-ketoglutaric acid
was stressed by Martius (1937) as being of special importance
because it can be a mother substance for important amino acids.
VON Euler, Adler, GUnther and Das (1938) could show that
a-ketoglutaric acid is able under physiological conditions to
bind ammonia and form glutamic acid in presence of the spec-
ific enzymes. This reaction is so much the more important as
the amino group of the glutamic acid can be transferred by

enzyme action to other a-keto acids (Braunstein and Kritz-
mann, 1937) and the a-ketoglutaric acid then reformed can
afterwards take up fresh ammonia. We can thus speak of a
catalysis of the linking of free ammonium, with formation of
amino acids. Adler, v. Euler, GOnther and Pless (1939)
consider it probable that direct coupling of the Ci breakdown
with the synthesis of glutamic acid is the most effective path
for the formation of amino acids. Ci provides not only the
substrate, a-ketoglutaric acid, but also the hydrogen for the
synthesis of glutamic acid. Since both iso-citricodehydrogenase
and glutamicodehydrogenase are present in all animal tissues,
these authors hold it probable that the process is a universal
cellular reaction in the body. In favour of this function o i
argues especially the fact that the breaking down of Ci occurs
chiefly in the renal parenchyma, which occupies a cen ra
position in amino-acid metabolism. The continua supply of
Ci to the kidneys would appear to be also quantitatively su
cient to have some effect; calculating on the basis of con-
ditions in man, about 10 grams would be
the kidneys. This suffices perhaps for the formation of t
greater part of the so-called dispensable ammo
may be absent in the food, as the body itself is
them up from other material. The questmn as to -h this
connexion between amino acids and Ci can be experimentally

demonstrated in vivo is dealt with in ^fomiation of

According to Krebs and Cohen (1939)

great importance in other tissues than
muscle. These tissues according to my ^ f

ally a high Ci-concentration, which may als p
of this connexion between Ci and ammo acids.

JOHAN mARTENSSON

Summary.

Following a review of previous investigations on the citric-
acid metabolism in animal bodies, an account is given of a
series of experiments which have yielded the following results:

The citric-acid content of blood serum (C/s) is considerably
higher than that of the red blood corpuscles (probably five or
six times higher).

Intravenously injected citrate (Ci) is rapidly eliminated from
the blood (up to 100 mgm per kg disappear almost completely
in 1 Va hours). The fall in concentration is the more rapid the
higher the serum Gi-ievel is.

The liver takes only a ver}" small part in this Ci-elimination,
which is shown by perfusion of the isolated liver, by Ci-injec-
tion direct into the portal vein, and by analysis of joint samples
from the portal and hepatic veins.

It is confirmed that orally administered Ci is rapidly ab-
sorbed. Even under fasting conditions, however, the C/s is
higher in the portal than in the arterial blood, which is con-
sidered to be due to formation of Ci during metabolic processes
in the intestine or through bacterial action; experiments
show the latter to be operative in rabbits.

After a functional cutting-off of the liver and portal area,
the C/s falls in the rabbit (owing to the prevented Ci-absorplion
from the intestine outweighing the power of the liver to break
down Ci) but remains fairly unchanged in the cat, and the body
is still able to remove intravenously injected Ci from the blood.

Ci is formed in the muscles, for the C/s is found to rise in an
eviscerated preparation as well as in a perfused, isolated hind-
leg preparation, and the venous blood coming from muscle
has a higher Ci-conccntration than the arterial.

After intravenous administration of Ci the urinary excretion
of Ci is very greatly elevated so long as the C/s is high, but
diminishes again as rapidly as the C/s falls.

Renal tissue has a great capacity to oxidize Ci, which is
evidenced b}'' perfusion experiments on isolated kidney. The
Ci-concentration in the renal-vein blood is 20 — 30 % lower than
in the arterial blood, even when the amount of Ci in the urine
is very small. After nephrectomy there is a rise in the C/s.

The Ci-concentration in liver and .skeletal muscle is lower
than in the serum, but in renal tissue and heart muscle it is
higher than in the serum. After administration of Ci the con-

ON THE CITRIC ACID METABOLISM IN MAMMALS 73

cenlration rises inconsiderably in liver and muscle, whereas in
renal tissue the rise is very substantial, which is regarded as
proof that Gi is unable, or only able with difficulty, to penetrate
into intact liver and muscular tissue. This explains the lack
of agreement between in-vivo and in-vitro experiments respect-
ing tlie oxidation of Ci.

The results of the experiments are discussed with reference
to the more recent conception of the formation, breakdown and
function of Ci in the body. It is stressed that the hitherto
published in-vivo experiments on precursors to the endogenous
Ci are not conclusive, since their results may depend solely on
a displaced proportion between the Ci-oxidation in the renal
parenchyma and the Ci-excretion. (Nor can the3% therefore,
be regarded as a corroboration of the citric-acid-cycle theory of
Krebs). The same applies to experiments on the influence of
food on the endogenous formation of Ci.

The fact demonstrated here that Ci is continually being
supplied to the kidneys and oxidized there suggests that Ci may
have a function to fulfil within the amino-acid metabolism.

CHAPTER III.

Wliat is the Cause of the HjTpercitricaemia
in Lesions of the Hepatic Parenchyma?

T hunberg (1933 a), on investigating the C/s in various
pathologic states, found a raised value in, among other
disorders, hepatitis. Oral administration of Ci to a hepatitic
subject was followed by a powerful rise of the C/s with a slow
return to the initial value (Thunberg, 1933 b). Tbe extensive
clinical studies of SjOstrOm (1937) showed very clearly that
the C/s rises so regularly in lesions of the hepatic parenchyma
that determination of this value affords a very valuable aid
to the differential diagnosis of hepatitis and obstructive
jaundice.

This h.'is subsequently been confirmed in clinical practice, in which the
Ci-test has come into extensive use, at any rate in the Scandinavian coun-
tries. The C/s determination seems to have gained still greater value for
differential diagnosis since it was combined with determination of the
serum-phosphatase activity by a simple and reliable method (Buen and
Buch, 1939), which renders possible a positive demonstration of an ob-
struction in the bile passages. But this latter lest is also positive in a
number of hepatitic cases (about 10 % according to Gutman, Olson,
Gutman and Flood, 1940), and hence it is just the combination of the two
tests that is so valuable.

From his animal experiments on these problems SjOstrOm
came to the conclusion that the amount of Ci-dehydrogenase in
the liver is reduced in lesion of the hepatic cells, with the re-
sult that the normal capacity of the liver to break down Ci
is diminished, this being considered the cause of the C/s rise.
In clinical literature (Buch, 1940) this explanation has been
formulated thus: The Ci-dehydrogenase content declines con-
siderably in parenchymatous diseases of the liver and there-
fore the Ci is not broken down in the usual manner but passes
into the blood to a greater or less extent.

ON THE CITRIC ACID METABOLISM IN MAMMALS

This very natural explanation cannot, however, be correct.
Those perfusion tests in which SjOstrOm found an immense
decomposition of Ci in the liver have proved inexact and their
results do not fit in with other known facts respecting the Ci-
melabolism. SjOstrOm was able to show that the C/s rises in
hepatic injur\’^ produced by allyl formiate, but, obviously, this
says nothing about the underlying mechanism. The perfusion
tests on experimentally injured liver, which SjOstrOm has also
taken as a support for his conclusion, cannot be awarded any
evidential value (even if the technique employed had been
correct), for, in two reported experiments with heparinated
blood, the Gi-melabolism was in one case only 1.5 %, in the
other 2.7 % , lower than in an uninjured liver. As a matter of fact,
rather would a higher Ci-nietabolisni have been expected in the
injured liver, since, as SjOstrOm asserts elsewhere in his work,
the necrotic liver contains large amounts of enzyme, which
should have a greater chance of coming into contact with its

substrate than in normal liver.

Experiments of mine submitted earlier in this work have
shown that in the rabbit and cat the kidneys play a consider-
ably greater part than the liver in the Ci-breakdown, and it
is evidently the kidneys which keep the C/s at a fairly constimt
value. Experiments on dogs and rats indicate the same condi-
tion there. It may accordingly be regarded as very probable
that the Ci-metabolism in man behaves m the same way.
Judging from animal experiments the renal oxidation o
rises with rising C/s, at least to begin with, so that the P^rcentag
difference between the Ci-concentrations in the f^tenal and the
renal-vein blood remains about the same. If it is
this also applies to man, then an assumption ^hat the^^

citricaemia in hepatitis (which may have a pyer

Ci per cc) depends upon a cessation of ^-oxidation ^
would presuppose that this Ci-break own i
eral times plater than that in the kidneys. This is most

^”?“tion, therefore, can scarce^^^ed «
than that the influence of the

the C/s is Mrecf. This is also suggested by

elevation of the C/s occurs ear y contemplated. Ei-

account for the rise two jg^nted under the in-

ther the Ci-formation m the > mpinbolism that a grave
fluence of the profound ehange m metabolism

JOHAN mArTENSSON

liver injury involves, or else this injury brings about an in-
hibition of the Ci-breakdown in the kidneys. Experiments with
administration of Ci to hepatitics speak decidedly in favour of
its being the Gi-oxidation that is reduced. A strongly retarded
return of the C/s to its initial value ought not to occur if the
hypercilricaemia depends upon an increased Ci-formation with
an unimpaired Ci-decomposing capacity. If the renal oxidation
of Ci is inhibited in some way or other, this is evidently com-
pensated for by an elevation of the C/s to such a level tliat the
Ci-breakdown becomes equal to normal when this new position
of equilibrium has been established. That is to say, the abso-
lute difference between the Ci-concentrations in the arterial and
renal-vein blood is equal to normal, though if calculated as a
percentage it is lower. In this connexion it may be mentioned
that the absolute increase of Ci in the circulating blood when
there is a rise of, say, 10 yjcc in man, which is regarded as
clearly pathologic, is not greater than about 0.o5 gram, thus very
low in relation to the power of the kidneys to break down Ci.

The breakdown of Ci in the renal parenchyma may no doubt
be inhibited in several ways as a result of a severe injury to’ the
liver-cells: (1) by accumulation of metabolic products which do
not undergo further decomposition in a normal manner, (2) by
the inability of the liver to deal with the substances absorbed
from the intestine, or by such substances entering to a greater
extent direct into the circulation, for instance, in cirrhotic condi-
tions, or (3) by toxic products from the injured liver paren-
chyma itself. Clinicians speak of a “hepatonephric syndrome”:
the kidneys have to take over the detoxicating function of the
body if the liver fails, but they do not possess the great margin
of safety of the liver and fall short of the needs rather soon
themselves. The toxic substances accumulate in the epithelium
of the convoluted tubules, where they may cause considerable
damage. In addition, a direct action on the renal parenchyma
by the same toxic agent as injures the liver must be taken
into consideration.

Without going into the clinical literature or other experi-
mental works on the functional connexion between liver and
kidneys, I am submitting here an account of a series of animal
experiments that may throw some light on the mechanism un-
derlj'ing the rise of the C/s in injurj’ of the hepatic cells. The
experiments are grouped on the same lines as those reported
in the preceding paragraph.

ON THE CITRIC ACID METABOLISM IN MAMMALS

The Oi metabolism after functionally cutting off the
liver and preventing absorption from the portal region.

In Chapter II (p. 48) an account is given of experiments
devised to arrive at the Ci conditions after functional elimina-
tion of the liver and portal area. The C/s was found to fall
considerably in the rabbit hut to keep fairly constant in the
cat. In- the rabbit the Ci-breakdown after intravenous admi-
nistration seemed to be fairly unchanged, A tolerance-curve
(i'ig. 9) shows, however, that the C/s attains an equilibrium
at a rallier high level, which is indicative of a certain amount
of inhibition of the Ci breakdown in the kidneys, although
this inhibition, in view of the rapid fall of the curve at the
beginning, cannot be large.

In the following experiment on a rabbit, in which the liver and portal
area were cut off from the circulation, the C/s difference between the
arterial and rcnal-vcin blood was only 14 % after 2 hours and 20 minutes,
thus indicating a distinct inhibition of Ci breakdown. In all experiments
of this type the C/s also shows a tendency to rise towards the end.

5. 4. 40. Rabbit, 2.7 kg. Urethane 1.25 gm/kg. Ligation of the arteries
to the portal area as well as of the portal vein. Continuous injection of
10 % glucose, !■ cc in 10 minutes.

Normal value CO 120 140 min. after ligation

C/s 71.9 27.8 28.4 28.7 y/cc.

Corresponding experiments on the caf show no fall but, if
anything, a small rise of the C/s (Fig. 9). ns may
terpreled as indicating that in this animal ^ ^

from the intestine and the Ci consumption in the 1 y^rjteep
in equilibrium with each other, but it may a so e
fact that in the cat there sets in a considerably earlier
effective inhibition of the renal ^i-breakdown ban m
rabbit. Such an inhibition appears very clearly after additi
of Ci to a cat prepared in this way. . , „

26. 5. 58. Cal, 3.1 kg. Urethane 1.25 gm/kg. experiment

to the portal area as well as of the portal vmu. “e jxp

. of „oeo„ 0d„,n„.ercd.

CM .

CO, a

Cx, 60 mglkg,

JOHAN MARTENSSON

Before Ci-iiij. 3 15 35 60 100 min. after Ci-inj.

C/s 42.2 854 510 399 321 282 y/cc.

A comparison of the above elimination curve with one ob-
tained from an intact cat (in which the kidneys also play the
essential role in the Ci-oxidation), or with one giving the Ci
breakdown in isolated kidneys, will clearly show that the break-
down of Ci in the renal tissue must here have been subject
to a pronounced inhibition, since so moderate a dose of Ci,
even after 100 minutes, still held the C/s up at a level seven
times higher than the initial value.

The C/s in a sample taken from the renal vein at the same time as the
last sample was, however, only 13.C % lower than in the arterial blood,
which certainly implies an inhibition, though scarcely so pronounced as
might be expected in this case. Certain difficulties, however, arise in
drawing conclusions from the difference between the C/s of the arterial
and that of the renal-vein blood, unless the blood flow through the kidneys
per unit of time is simultaneously measured.

The Oi-metabbKsm after functionally cutting off the
liver but retaining the portal circulation intact.

The object of this type of experiment is to study tlie in-
fluence on the C/s of that part of a deficient liver-function
which consists in the liver not being able to deal with the sub-
stances coming from the intestine, or in these substances pas-
sing direct into the circulation, as in conditions of cirrhosis.

The preparation technique was as follows: In the right jugular vein
of the heparinated animal was tied a coarse, bent cannula, which was
connected by a rubber lube with a cannula fitted into the portal vein.
The system Avas filled with Tyrode’s solution. The portal vein was ligated
as closely as possible to the liver, then clamped peripherally, Avhereupon
the cannula was quickly tied in and the circulation could immediately start
in the anastomosis. Then the hepatic artery was ligated. In spite of
continuous injection of glucose (150—200 mgm per kg per hour) the
animal did not .survive more than 1 ‘/i — 2 hours.

In these experiments with intact portal circulation, however,
there is a vigorous rise in the Cfs also in the rabbit. This must
mean that substances absorbed from the intestine cause a fur-
ther inhibition in the breakdown of Ci if they enter the cii’-
culation -without first being dealt with by the liver. For the
preceding type of experiment clearly shows that an increased

ON THE CITRIC ACID METABOLISM IN MAMMALS

Ci-formation is not established if the liver is put out of function.
The C/s rise seems to be less in animals that have been fasting
rather a long time, i. e. in which the absorption of substances
from the intestine cannot be very large.

The follo^ving experiments may be adduced:

2. 5. 38. Rabbit, 2.7 kg. Starved less than 24 hours.

Normal value 20 40 67 min. after op.

C/s 109 140 158 170 r/cc

5. 4. 40. Rabbit, 3 kg. Starved 24 hours.

Normal value 60 min after op.

G/s 61.7 107 y/cc

The rabbit died as a result of pulm. embolus after another 40 min. and
before further samples had been withdrawn.

13. 11. 39. Rabbit, 3.9 kg. Starved 48 hours.

Normal value 30 60 90 140 min. after op.

C/s 61.9 78.5 84.5 97.3 112 y/cc

20. 11. 39. Cat, 2.9 kg. Starved 24 hours.

Normal value 30 65 min. after op.

C/s 50.7 67.6 94.6 y/cc

G. 5. 40. Cat, 2.7 kg. Starved 72 hours.

Normal value 30 60 90 min. after op.

C/s 46.4 47.8 51.2 64.8 y/cc

In this case the rise of the C/s was not much greater than when the
liver was cut off from the general circulation and the portal flow was
arrested, but here there was obviously nothing that could be absorbed
from the empty gastro-intestinal canal.

When Ci is administered to a cat provided with a portal vein-
jugular vein anastomosis, the Ci-elimination is retarded an
soon ceases at a high level, which shows that the breaking
down mechanism is effectively arrested.

7.- 5. 40. Cat, 2.3 kg. Starved 24 hours. After the anastomotic circula--
tion had been in action 20 min., Ci equivalent to 60 mgm per 'g
injected for exactly five minutes.

Before inj. 3
C/s 45.5 622

358

35 60 90, min. after inj.

267 270 266 y/cc

> • • • • • rtO.U

Boothby and Adams (1934) investigated the urinary

logs following removal of the liver. After t ^ internret

:retion rose considerably above the normal, which ^

IS indicating that possibly Ci in the dog is normally utilized or destroyed

JOHAN mArTENSSON

by the liver. Instead of this, the rise probably depends on an arrest of
the normal Ci-breakdown in the renal parenchyma following the hepat-
ectomy, the amount of Ci in the urine. being thereby increased. This in-
crease stands out especially clearly in the dog, as the normal Ci-con-
cenlration in the urine of this animal is e.xtremely low.

The Oi-metabolism in experimental lesion of the hver.

SjOstrOm’s work (1937) contains a survey of the technique
for producing a liver injury by chemical means. In a couple
of cases I duplicated Sj5str5m’s experiments, administering
allijl formiate (Allylium formicicum, Schering-Kahlbaum) in-
iraperitoneally to a rabbit. Macroscopically the liver there-
upon presented a picture corresponding well with that des-
cribed by SJiisTRCiM from his experiments, and the C/s also
rose moderately at the end. However, it is doubtless difficult
in such experiments to keep the experimental conditions con-
stant for any long period of time. In the raljbit the C/s varies
rather much w'ith the food, and, even if the latter is kept con-
stant, the C/s drops considerably if the animal under the in-
fluence of the intoxication refuses to eat. Moreover, intra-
peritoneal injection of allyl formiate appears likely to cause an
irritation leading to peritonitis, and such inflammatory states
depress the C/s. It may be these factors that caused the in-
cipient decline of the C/s which SjOstrOm found in all his ex-
periments, rather than an assumed increase of the Ci-dehydro-
genase activity in the injured liver.

Anj'^ attempt to produce injury of the liver chemical means
will always involve the risk of directly affecting the renal
parenchyma. In the experiments with allyl formiate, the kid-
neys appeared to be intact (at all events macroscopically). In
an experiment with chloroform poisoning, in which O.i cc of
chloroform per kg was daily injected subcutaneously, the rab-
bit was killed after eight days, and the kidneys then exhibited
the same pronounced fatty infiltration as the liver. The serum
samples were then markedly icteritious. The C/s behaved as
follows:

4.4. G.4. 7.4. 8.4. 9.4. 11.4

C/s .... 107 96.3 122 105 124 130 y/cc.

A sample of the renal-vein blood, whicH was taken at the same time as the last
sample from the carotid artery, had a value of 123 per cc. The small
difference as against the arterial value points to a pronounced arrest of

ON THE CITRIC ACID METABOLISM IN MAMMALS

Ihe Ci-brcakdown, although this may here he due to the direct chloroform
action on the renal parenchyma. Chloroform does not, accordinglj’, give
pure experimental conditions.

I therefore tried to produce a servere hepatic lesion without
administering any toxic substance. With this end in view I
utilized the condition demonstrated by MoMichael (1937) that
the rabbit’s liver obtains almost all its oxygen from the hepatic
artery and that necrosis of the liver sets in if the hepatic artefy
is ligatured.

The following experiment shows how the C/s varies when the
hepatic artery is ligated under ether narcosis and the rabbit
is afterwards allowed to revive.

C. 12 . 39 . Rahhil, 5.1 kg. Lively just after operation, began to be
.sluggish 17 — 18 hours later, more and more so subsequentlj'. Died after
28 hours.

Normal value 7 17 24 28 hours after lig.

C/s G.3.0 G2.0 50.4 G7.G 107 y/cc.

First comes the usual post-operative decline in the C/s. This counteracts
a possible rise due to the hepatic injury, svhich rise, however, makes its
appearance at the end. Macroscopically the liver seemed to be necrotic
throughout.

Seeing that the post-operative reaction disturbs the experi-
mental conditions, I went over to allowing the animals to he
under light narcosis the whole time. Then there occurred t e
•same thing as was oliserved in the experiments with nephrec-
tomy: the course was run much quicker under narcosis, e
animals dying in five or six hours after ligation of the hepatic
artery. During this time the C/s rose evenly up to double the
value or higher.

The following cxpcrimcnl is submiUed as typical: 4. 4. 40. Rabbit,
2,0 kg. Urethane 1.25 gm/kg.

Normal value 3 4 5 V 2 hours after lig.

Q/s G0.7 85,2 99.8 115 y/cc.

The blood pressure was somewhat inferior in the sample last taken.
The rabbit died soon after withdrawal of this sample.

In Fig. 15 curve III gives the Ci-elimination ™ ^

eriment in which Ci was administered
epatic artery. Ci was injected two “““

as appiied, and after a further 2 % hours the
o strong effects from its liver injury. In spite of this, the

JOHAN MARTENSSON

Ci-elhninatioii is very much retarded. For comparison I am
submitting a cun’e showing the elimination after addition of
the same amount of Ci to a normal rabbit (Curve I) and to
a rabbit with its liver entirely cut off from the general circula-
tion (Curve 11). The experiment shows that for breaking down
Ci in the body an injured liver is inferior to no liver at all,
a further support for the view that the influence exercised by
hepatic lesion on the Ci-metabolism is indirect.

Broadly viewed, these different types of experiments on
functionally eliminated livers or on experimental liver injuries
point in the same direction as regards the influence on the C/s.
It is evident, however, that the retardation of the Ci-breakdown
in the kidneys is greater when an abundant ab.sorption from
the intestine is proceeding while the liver is out of function, or,
if the liver is injured, while it is in communication with the cir-
culation and toxic products can flow out. The Ci-elimination
curves show a perfect parallel to what I found when urethane
M'as given for narcosis intravenously instead of subcutaneously
/MArtensson, 1938). Then the substance evidently accumu-
lates in the renal parenchyma to so high a concentration that
it directly inhibits the Ci-dehydrogenase through its narcotic
effect.

Half an hour after intravenous injection of 25 % urethane to an amount
of 1.5 gm per kg in a rabbit, the C/s %vns 62.8 7 in the arterial blood and
58.5 y in the renal-vein blood. The difference amounts to only 7 % of the
arterial value, which shows that the renal oxidation of Ci is greatly
reduced.

Is the Oi-metabolism in vivo connected with the amino-

acid metabolism?

As was previously mentioned, I have been able to show that
administration of malic acid inhibits the breakdown of Ci in
the kidne 3 ^s (MArtensson, 1939). Possiblj" the same mecha-
nism is behind the effect of all those substances which, on
Ijeing added, have proved able to increase the Ci-excretion in
the urine. Whether any of these metabolites contribute to
elevating the C/s in injurj' of the hepatic cells is an open question
until we have a better knowledge of their metabolic condi-
tions in the bodj'.

Another reaction suggests itself more readih' if the cause of
the C/s rise in hepatic. lesion is sought in a .specific inhibition,
viz. glutamic-acid formation from Ci via a-ketoglutaric acid.

ON THE CITRIC ACID METABOLISM IN MAMMALS

which according to Adler and others (1939) is the most effec-
tive route for formation of amino-acids in the body. The
Ijreakdown of amino-acids seems to take place chiefly in the
liver, and the latter’s capacity to do this is reduced when this
organ is injured. .After intravenous injection of amino-acids
(7 mgm amino-acid nitrogen per kg) to rabbits, the amino-

c/s r/

cc.

No operation. 00 mg dlJ^g-

JNO operaiiuii.

Ligation of arteries to portal area and of portal vein.

Ligation of hepatic artery. 2 hours later 00 mg dlkg
Same operation as II. 00 mg d and 100 mg gluiamtc

I. Rabbit, 2.1 kg.

II. Rabbit, 4.0 kg.

60 mg dlkg.

III. Rabbit, 4.1 kg.

IV. Rabbit, 3.0 kg.
acidikg.

acid content of the blood already returns to

after SO-CO minutes (Kitamuba, 1937) but if “

injured, espeeially by chloroform, the am.no-acd content rises

JOHAN mARTENSSON

and the tolerance curves assume an ascending character. The
elimination, however, is as rapid if the kidneys are extirpated.
This agrees with an investigation by Svedberg, Maddock and
Drury (1938): The amino-acid content remains normal after
nephrectomy but rises vigorously after hepatectomy. In the
course of necrotic processes in the liver, amino-acids are also
formed during tissue disintegration, and as a result a con-
siderable rise of the amino-acid level in the blood can be found
in pronounced liver injuries. In such cases a likely assumption
is that a displacement may occur of the equilibrium in the
reversible reaction between Ci and glutamic acid with, as its
result, a compensatory increase of the blood Ci in order to
restore the reaction equilibrium.

I have therefore conducted some experiments to learn how
addition of glutamic acid affects the in-vivo metabolism of Ci.
Orten and Snhth (1937) had found only a slight increase of
the Ci-concentration in the urine in dogs after intravenous
injection of glutamic acid. But this slight effect is no doubt
attributable to the very rapid breakdown of the amino-acids
by the liver and their disappearance from the blood. In agree-
ment with this, I found that the effect was slight after intra-
venous injection of glutamic acid (Kahlbaum), 100 mgm per
kg being administered to a rabbit (without narcosis).

12. 10. 39.

Normal value 10 .20 40 min. after inj.

C/s 71.4 75.2 79.C 81.8 y/cc

After joint Ci and glutamic-acid administration the Ci-elimi-
nation was however found to be distinctly retarded, especially
at first when the glutamic content of the blood could be as-
sumed to be high.

16. 10. 39. Babbit, 2.3 kg. No narcosis. Inj. of 100 mgm Ci per
kg -f 100 mgm glutamic acid per kg during exactly five minutes.

Normal value

86 min. after inj,

88.7

525

352

207

169 7/cc

(87.6

510

230

150

135 ., )

Tlie bracketed figures are corresponding values from a control test with
the same amount of Ci but without glutamic acid, likewise on a non-
narcotized rabbit.

In order to escape the rapid break-down of glutamic acid,
and to gain time for its effect to be exercised, I carried out
the same experiment on a rabbit ■whose liver and portal area

ON THE CITRIC ACID METABOLISM IN MAMMALS 85

had been functionally cut off (after this operation the amino-
acid nitrogen in the blood rises by 1 — 2 mgm per 100 cc in
two hours). After injection of 100 mgm of glutamic acid per
kg (Sodium glutamate, mono, B. D. H.) the amino-acid nitrogen
in the blood rose from 6.3 to ll.i mgm per 100 cc and after a
further 95 minutes was 12.3 mgm per 100 cc. The Ci-elimi-
nation curve after injection of 60 jngm of Ci per kg is given in
Fig 15. curve IV, which should be compared with curve II as
representing a control test without glutamic acid. As will be
seen, the retardation is considerable. Direct measurement of
the difference between the C/s of arterial and that of renal-
vein blood gave a value of 17.5 %, which implies a consider-
able inhibition.

Tlic amino-acid estimation was carried ont by Folin’s colorimetric
method (Folin, 1922) with unlakcd blood as starting-point according to
Folin (1930). As standard I used a solution of glycine (Amino-acetic acid
Analar, B. D. H.), and tlie amino-acid reagent was likewise from the
B. D. H.

The amino-acid estimation in the investigations referred to above was
made by the same method, and hence the values are comparable. The
average value of the amino-acid nitrogen for the rabbit is 7.9 mgm per
100 cc according to Kitamuba (1937), 8.1 mgm per 100 cc according to
SvEDBERG, Maddock and Drury (1938).

Lastly, an experiment is submitted in which glutamic acid
was added during perfusion of isolated cal kidneys.

IG. 11.39. Cat, 3.2 kg, wt. of kidney 31 gm. Perfusion fluid 220 cc
heparinated blood with almost 20 % Tj'rode’s sol. Haematocrit reading
34 %. Min. vol. between 48 and 58 cc. Pressure increasing from 80 — 110
mm Hg. Urine secretion very slight during whole experiment. After a
period for stabilization, Ci was added in continuous drip, 40 mgm during
20 min. Glutamic acid (neutralized), altogether 500 mgm, was then added,
whereupon another 40 ■ mgm of Ci was supplied during 20 min. The
calculation of the Ci-consumption was effected as in previous perfusion
tests. Blood vol. computed during both periods at 215 cc, the Gi-content of
the perfusion blood in relation to the plasma in the first period at 75 %,
in the second at 80 %, as a certain volume of fluid was added.

C/s Estimated Ci-consumption

1st period 25.7 — 160 yiee 18.4 mgm

2nd 64.4—224 „ 12.5 „

The breakdown of Ci in the kidneys thus lell, after addi-
tion of glutamic acid, by about '30 % , although the rise of the
average C/s in the later period, from 93 to 144 yfee, ought to
have increased the' oxidation by 50 %. Judging from these

JOHAN MARTENSSON

experiments there thus exists a rather high probabilitij that the
reaction between Ci and glutamic acid can also play a role in
vivo.

Even if the amino-acids are not broken down in these later experiments,
the possibility must be conceived of the glutamic acid being changed by
the transamination that is constantl}- taking place (Schoenheimer, Ratner
and Rittenberg, 1939). In the course of this, it is thought that alanine
or aspartic acid is first formed (Cohen*, 1939) and the glutamic acid is
converted into n-kctoglutaric acid, w’hich as a breakdown product of Ci
ought to inhibit the further disintegration of this acid. Measurement of
the amino-acid nitrogen in the blood should give a total picture of what
happens even if a transamination takes place. However, I have not yet
had an opportunity of investigating whether an amino-acid formation
actually occurs during kidney perfusion with addition of Ci in such amount
that it is able to give a clear verdict. In a preliminary experiment with
anastomosis between the renal and jugular veins, the amino-acid con-
centration was certainly higher in the renal-vein blood after addition of
Ci (8.4 mgm per 100 cc as against 7.9 mgm per 100 cc in the arterial blood),
but the point was not closely investigated.

The following account shows how the C/s and the amino--
acid content of the blood are changed after the different opera-i
tions dealt with in this chapter. The related estimations of the
C/s and the amino-acid nitrogen were made on the same blood
samples.

1) Functional cutting-off of liver and portal area:

Rabbit: C/s

71.9-

-27.3 — 28.7 y/cc

Amino N ...

7.4

— 8.5 mg p. 100 cc

Cat; C/s

31.0

— 42.2 y/cc

Amino N ...

5.9

— 8.7 mg p. 100 cc

2) Functional cutting-off of

liver

but intact portal circulation:

Rabbit: C/s

G1.7

— 107 y/cc

Amino N ...

6.8

— 14.3 mg p. 100 cc

Cat: C/s

4G.4

— G4.3 y/cc

Amino N

5.4

— 7.2 mg p. 100 cc

3) Ligation of hepatic artery

■:

Rabbit: C/s

60.7

— 115 y/cc

Amino N ...

7.3

— 11 mg p. 100 cc

Rabbit: C/s

59.C

— 141 y/cc

Amino N ...

— 10.8 mg p. 100 cc

A definite parallel is thus noticeable between the degree of
the rise in the amino-acid content and of that, in the C/s, and
hence it is possible that there is also a causal relation between
them. In this way these experimental observations might lead

ON THE CITRIC ACID METABOLISM IN MAMMALS

to the following view as to the reason for the applicability
of the G/s determination as a liver-function test; The break-
down of Ci takes place chiefly in the renal parenchyma. The
Ci-metabolism is, however, intimately associated with proces-
ses within the carbohydrate and protein metabolism, especially
the transformation of amino-acids, which processes have the
closest connexion with the metabolic function of the liver.
Since Ci is a most readily convertible substance and occurs
in very low concentration, it responds verj’^ sensitively to va-
riations in the quantitative conditions of the reactions in which
it takes part. For this reason we can employ the C/s as a
sensitive indicator to the functional state of the liver, as we
possess means of exactly registering even minor variations in
the C/s.

Summary.

Estimation of the serum citric acid has proved a valuable
aid for the differential diagnosis of hepatitis and obstructive
jaundice. The hypercitricaemia attending injuries of the hep-
atic parenchyma cannot, however, according to experiments
submitted in Chapter II, be due to a diminished Ci-oxidation in
the liver, as was previously assumed.

Animal experiments show that if the liver is put out of
function, or is injured, there occurs an inhibition of the Ci-
oxidation in the kidneys, with a diminution of the difference
between the C/s of the arterial blood and that of the renal-vein
blood and a markedly retarded elimination of the added Ci.

Administration of glutamic acid retards the Ci-elimination
and diminishes the Ci-oxidation in the kidneys in perfusion
experiments. In those experiments in which the liver is put
out of function, or is injured, there is a definite parallel be-
tween the rise in the C/s and that in the ammo-acid content of
the blood. Therefore, as a possible specific cause of the hyperci-
tricaemia in injuries of the hepatic parenchyma, it is suggested
that a deterioration of the amino-acid breakdown in the liver
may bring about a displacement of the equilibrium in the
reaction between the Ci-breakdown and the glutamic-acid for-
mation, causing, a rise in the C/s. Ci, being a substance .that is
very readily oxidized and that occurs in low concentration,
responds very sensitively to every change in the reaction en-
vironment.

CHAPTER IV.

Some Experiments on tlie Ci-metaboUsm
attending Changes in the Acid-base
Equilibrium.

E arlier in this work it was mentioned on a couple of oc-
casions that changes in the acid-base equilibrium may
bring about alterations in the C/s and affect the experimental
results. For this- reason I am submitting below a short review
of the literature on this question as well as some animal ex-
periments, although I have not yet been able to make a de-
tailed experimental inquiry' into the problems involved.

OsTBERG’s investigation (1931) shows that the Ci-excretion in man is
always diminished after administration of acids or of agents producing
acidosis as well as after carbohydrate starvation, while it is always in-
creased after administration of alkalies. Ostberg points out that this
Ci-cxcrclion in the presence of excess of alkalies acts sparingly on the
"fixed” anions and can therefore play a part in the renal regulation of
the acid-base conditions. At that time no definite opinion could be ad-
vanced as to the source of this Ci or as to the cause of the variations
induced by changes in the acid-base equilibrium. A later investigation led
O.sTBERG (19341 to the conclusion that the Ci-excretorj' changes induced by ad-
ministration of acids or alkalies cannot be governed by changes in the C/s,
since no sure correlation could be established betw’een them.

Kuyper and Mattill (1933) confirmed Ostberg’s results. They were
able to show that, in the rabbit, inanition causes the C/s to fall but that
simultaneous administration of bicarbonate counteracts this fall. If am-
monium chloride is supplied, the C/s drops considerably, though the
authors ascribe most of the effect to the inanition. Alkalosis through
hyperpnoea increases the Ci-excretion, powerful bodily movement diminishes
it. The authors state that this diminution is evidently not a consequence
of a decline in the C/s and that ‘‘there is as yet no information ns to what
the nature of the kidney threshold for citrate may be”.

Ostberg’s results were further confirmed by Boothby and Adams

ON THE CITRIC ACID METABOLISM IN MAMMALS 89

(1934) who point out, however, that on the basis of the clinical material
(hey have been unable to discover any correlation that would be of clinical
significance.

Lenn6r (1934) reported a couple of experiments on man, according to
which there is possibly a rise of the C/s after oral administration of
bicarbonate, a fall after' ammonium chloride.

Sherman, Mendel and Smith (1936 a) likewise confirm that the urinary
Ci varies in amount direct with the pH of the urine, irrespective of the
cause of the pH variations. The increase in the Ci-excretion following
intake of bicarbonate is especially marked (up to 100 times) in dogs and
rats with their normally very low Ci-concenlration in the urine. Tlie
authors infer that the increased Ci-excretion is not caused by an increase
of the C/s but must be due to a specific activity of the kidneys.

This remarkable phenomenon in Ci-excretion associated
with change in the acid-base equilibrium, and the specific renal
activit}’^ underlying it, can now be readily 'explained, since my
experiments have shown that Ci is being constantly supplied
in relatively large (juantities to the kidneys and there oxidized,
while in normal cases only a small proportion of it is excreted.
The Ci which after administration of alkali is excreted in in-
creased amount is thus readily available on the spot. A re?
arrangement of the renal activity in the service of the acid-
base regulation may involve a change in the relative proportion
of Ci-breakdown and Ci-excretion without necessarily altering
the total elimination of Ci in the kidneys.

Some experiments on rabbits, however, showed that an aci^-
dosis produced by ammonium chloride or calcium chloride can
depress the Cis very considerably.

14. 11. 39. Rabbit, 4 kg. Administration through stomach-sound of
12 cc of 25 % ammonium chloride = 0,75 gmfkg.

Normal value 3 Vs 6 */• 22 31 55 hrs after admin.

C/s 68.2 52.2 41 26.6 28.5 47.9 y/co

In this case inanition cannot play any rdlc, for the rabbit had already
starved 24 hours before the beginning of the experiment and was given
food again 22 hours after the ammonium chloride was supplied. In spite pf
this the C/s kept Jo'w for a long time. (The pH of the urine, tested with
Lyphan paper, dropped to 5.3).

SO. 11. 39. Rabbit, 4 kg. Administration through stomach-sound of 30 cc
of 10 % ammonium chloride = 0.75 gmikg.

Normal value 5 Vs 16 Vs 20 hrs after admin.

C/s 90.1 44.9 27.9 26 7 /cc

90 JOHAN MARTENSSON

The urinnrj’ Ci-concentralion fell from 167 y/cc (pH 8.5) to 31.8 y/cc
(pH 5..^).

Jl. 12. 3D. Rabbit, 4 kg. Administration through stomach-sound of
40 cc of 10 % CaCh — 1 gmikg.

Normal value 22 38 74 hrs after admin.

C/s 33.7 19 25.1 51.1 y/cc

The urinarj' Ci-concentration dropped from 71.9 y/cc (pH 8.3) to 29 y/cc
(pH 5.2). On autopsy a few days later it was found that the rabbit had
asciles and swollen lymph-glands in the abdomen, this being perhaps the
cause of the low initial value.

After oral administration of sodium bicarbonate to a rabbit
no distinct elevation of the C/s was observed.

21. 11. 39. Rabbit, 2.7 kg. Administration through stomach-sound of
20 cc of 10 % sodium bicarbonate 0.75 gmjkg.

Normal value 2 ‘/* 6 12 hrs after admin.

C/s 93.5 97.8 99.7 92.6 y/cc

At the beginning of the experiment the urine was already so alkaline
(pH 8.4) that no further rise of the pH ensued. The urinary Ci-con-
centration rose from 176 to 305 y/cc.

On the other hand, the elimination of Ci from the blood was
distinctly retarded when Ci was given at the same time as sodium
bicarbonate was infused intravenously, although the Ci-excre-
tion in the urine was considerably larger than normally. Si-
multaneous measurements of the C/s in the arterial and renal-
vein blood showed a clear retardation of Ci-breakdown in the
kidneys.

12. 5. 40. Rabbit, 5 kg. Urethane 1.25 gm/kg. Continuous intravenous
injection of sodium bicarbonate, 0.35 gm/kg per hour. After administra-
tion of 100 mg Ci per kg the C/s was as follows:

Normal value 3 15 35 60 90 120 min. after inj.

C/s 80.5 585 380 282 183 138 120 y/cc.

During the experiment 93 mgm of Ci, or 18.6 % -of the added amount,
passed away with the urine (Cf. experiment in Fig. 11). At the last
sampling the C/s of the renal-vein blood was only 15.8 % lower than that
of the arterial blood.

In a couple of preliminary experiments on perfused isolated
cat-kidneys I have studied the changes that take place in the
Ci-melabolism when the reaction in the perfusion fluid is dis-
placed by changes in the carbon-dioxide content of the gas-

ox THE CITRIC ACID METABOLISM IN MAMMALS

mixture bubbling throiigli. This procedure is the only safe one
if the intracellular hydrogen-ion-concentration is also to be
affected.

Ahlgren (1930) demonstrated the importance it has for experiments on
suHidving organs to maintain the bicarbonatc-COs buffer system in the
scrum Saline solution: if air or pure oxygen is passed through a serum
saline solutioi! containing bicarbonate, the pH is rapidly displaced to the
alkaline side. A displacement of this kind also occurs if blood is used as
an experimental medium, although it is there counteracted to some extent
by the great buffering capacity of the blood. In an experiment conducted
by Hasselbalch the carbon-dioxide tension was doubled without change
in the bicarbonate content, the pH of the total blood falling by 0.19 instead
of the theoretical O.so value.

My experiments were arranged in the following way: After a stabilizing
period, 30 mgm of Ci was added during 20 minutes, the “normal” gas-
mixture, Os + a % COj, being meanwhile bubbled through the perfusion
blood. Then the gas-mixture was altered to Oi -f 10 % COi, and after an
adjustment-period of ten minutes another 30 mgm of Ci was added during
20 minutes. Thereafter the Ci-addition was repeated once more, now with
a gas-mixture of O? + 1 % CO«. The consumption of Ci during the differ-
ent test-periods proved to be as follows:

Oi + 5 % COs Oj + 10 % COj Os + l% COs
17.2 mg 22.0 mg 19.5 mg Ci

In another c.xperimcnt this change of the buffer system was inversed.
In each test-period here 40 mgm of Ci were added. The Ci-consumption
during the different periods was:

Os -k 0 % CO; 0; + 1 % CO; Os "MO % CO;

29.7 mg 25.7 mg 29.4 mg Ci

According to these experiments the Ci-consumption in tlic
kidneys is increased by a displacement of the reaction in an
acid direction and diminished by a displacement in the alkaline
direction. The changes are not especially great, but may 'be
considered as certain. And a relatively small change in the Ci-
breakdown can in course of time alter the C/s very consider-
ably. For that quantity of Ci which is excreted in the urine a
small change of the oxidation may play a great role, as norm-
ally this part is a verj’^ small proportion of the total convex -
sion. According to the conception of the course of events, to
which these experiments give rise, the changes in the C/s and
the Ci-excretion are not directly dependent on each other but
both are determined by the intensity of the Ci-oxidation in the
renal parenchyma.

JOHAN MARTENSSON

In this field of inquirj-, however, there remain many questions un-
solved. It is liighly probable that the Ci-melabolism is bound up with the
amino-acid metabolism. The question may then be whether an increased
Ci-oxidation depends upon a supply of HsN, such as when ammonium
chloride is administered or when the ammoniac defences of the kidneys
are brought into play inwicidosis. Ostberg’s (1931) curves show with all
clearness that the amount of Ci in the urine is inversely proportional to
that of the ammoniac. Krebs and Cohen (1939) state that addition of
H«NC1 to renal cortex in in-vilro experiments markedly increases glutamic-
acid formation and respiration in presence of n-ketoglutaric acid or of
substances that can form this acid. I have found no distinct increase of
the Ci-climination after administration of H«NC1 direct intravenously in
rabbits, but the strong central .irritation associated with such probably
disturbs the experimental conditions, and this question should therefore
preferably be studied by perfusion tests on kidneys.

I have not closely studied the question of whether an acidosis also
brings about a diminished Ci-formation in the body. In an experiment in
which acidosis was induced with ammonium chloride, the C/s was 30.1 y
per cc in the arterial blood, 17.6 y in the renal-vein blood. This differ-
ence certainly implies a powerful Ci-oxidation in the kidneys, but absol-
utely it is nor so large that a Ci-formation equal to the normal can be
expected.

Nor has it been elucidated whether the Ci-metabolism in those pathologic
conditions in which the C/s is generally low (postoperative conditions,
thrombosis and inflammations) is associated with an acidosis, as has been
surmised, or with changes in the protein metabolism. Tlie postoperative
drop in the C/s must, in any case, also depend on something other than
an increased renal oxidation of Ci, since in the rabbit it also occurs after
bilateral nephrectomy (See Fig. 14). In two rabbits whose kidneys had
been extirpated, the C/s fell considerably towards the end, when the
animals exhibited signs of severe inflammation of the intestine.

The vigorous decline of the C/s after administration of salicylic acid
(.\i.wALL, 1938) does not appear, according to later investigations, to be
due to a direct salicylic action but to an acidosis produced by salicylic acid
(Al.w.\u. and MArtensson, unpublished).

Summary.

Earlier research-workers have shown that administration of
acids or of agents producing acidosis diminishes the Ci-ex-
cretion, while administration of alkali increases it. These
changes in the Ci-excretion could not he explained by cor-
responding falls or rises of the C/s, but were assumed to he
dtie to a specific renal activity. The situation is now amenable
to ready explanation, since my experiments have shown that

ON THE CITRIC ACID METABOLISM IN MAMNLVLS

the Ci normally excreted is merely a small proportion of that
oxidized in the renal tissue. Under the influence of a re-
arrangement of the renal activity in the service of the acid-
base regulation, a displacement of the proportion between Ci-
breakdown and Ci-excrelion can therefore occur without ne-
cessarily affecting the C/s.

An acidosis brought on by ammonium chloride or calcium
chloride can also depress the C/s very materially in the rabbit.
Adminislnttion of bicarbonate gives a doubtful rise of the C/s,
but retards the Ci-climination, the Ci-excrelion being then in-
croa.scd. The Ci-melabolism in the renal parenchyma, studied
by means of perfusion tests on Lsolaled cat-kidneys, rises in
response to* displacement of the reaction in an acid direction,
falls in response to displacement in an alkaline direction. It is
therefore possible that changes of the Ci-concentration in both
blood and urine are secondary phenomena to a change of the
Ci-oxidation in the renal parenchyma.

References.

(* indicates that the original article has not been available.)

Abdcn, N. O. and Gisselsson, L., Acta med. scand. 1938, Suppl. XC,
' ' 444.

Abraham, E. P. and Adler, E., Biochem. J. 1940, 34, 119.

.4dams, M., J. biol. Chem. 1931, 92, Proc.

Adler, E., Euler, H. v., GCnther, G. and Pless, M., Biochem. J.
1939, 33, 1028.

Ahlgren, G., Skand. Arch. Ph}’siol. 1930, 69, 1.

Alwall, N., Ibid. 1938, 80, 27.

Andersson, B., Z. physiol. Chem. 1933, 217, 186.

Andersson, B., Dissert. Stockholm 1938.

Annau, E., Banga, I., GOzsy, B., Huszak, St., Laei, K., Straub, B.

and Szent-GvOrgyi, A., Z. physiol. Chem. 1935, 236, 1.

Banga, I., Ochoa, S. and Peters, R. A., Biochem. J. 1939, 33, 1980.
Batelli, F. and Stern, L., Biochem. Z. 1911, 31, 478.

Beecher, H. K., The physiology of anesthesia. Oxford Univ. Press
1938.

Bernheim, F., Biochem. J. 1928, 22, 1178.

Blixenkrone-Moller, N., Skand. Arch. Physiol. 1937, 77, 15.
Blixenkrone-Moller, N.,Disputats, Copenhagen, 1938.

Boothby, W. M. and Adams, M., Am. J. Physiol. 1934, 107, 471.
Braunstein, a. E. and Kritzmann, M. G., Nature 1937, 140, 503.
Breusch, F. L., Z. physiol. Chem. 1937, 260, 262.

Breusch, F. L., Biochem. J. 1939, 33, 1757.

Buch, I. and BucH, H., Acta med. scand. 1939, Cl, 211.

Buch, H., Nord. Med., 1940, 5, 125.

Cohen, P. P., Biochem. J. 1939, 33, 1478.

Dann, W. j.. Ibid. 1931, 25, 177.

Davies, D. R. and Quastel, J. H., Ibid. 1932, 26, 1672.

Deuticke, H. j. and Hollmann, S., Z. physiol. Chem. 1939, 258, 160.
Dickens, F., Chem. and Ind. 1940, 69, 135.

Dixon, M., Ann. Rev. Biochem. 1939, VIII, 1.

Drury, D. D., J. exp. Med. 1929, 49, 745.

Euler, H. v., Adler, E., GOnther, G. and Das, N. B., Z. physiol.
Chem. 1938, 254, 61.

Euler, II. v. and Klussmann, E., Naturwissenschaftcn 1934,22, 777.
Euler, H. v. and Schlenk, F., Klin. Wschr. 1939, 18, 1109.

ON THE CITRIC ACID METABOLISM IN MAMMALS

Foun, O., J. biol. Chem. 1922, 51, 577; 1930, 86, 173.
forth, O. and Herrmann, H., Biochem, Z. 1935, 280, 448.
Gemmill, C. L., Skand. Arch. Physiol. 1934, 67, 201.

GrOnvall, H., Ada ophlalmolog. Suppl. 1937.

Gutman, A. B., Olson, K. B., Gutman, E. B. and Flood, C. A.,
.1. din. Invest. 1940, 19, 129.

Harrison, K., Biochem. J. 1939, 33, 1465.

Hathaway, M. L. and Meyer, F. L., .1. Nulrit. 1939, 17, 419*.
Himsworth, G., j. Physiol. 1938, 91, 413.

Hirschlaff-Lindgren, B., Skand. Arch. Phj'siol. 1937, 76, 15.
ISHIHARA, T., Arb. med. Fak. Okayama, 1938, 6, 461*.

Johnson, W. A., Biochem. J. 1939, 33, 1895.

Kitamura, I., Jap. J. Gastroenterol. 1937, 9, 166*.

Knoop, F. and Martius, C., Z. physiol. Chem. 1936, 242. 204.
Krebs, H. A., The Lancet, 1937, II, 736.

Krebs, H. A. and Cohen, P. P., Biochem. J. 1939, 33, 1895.

Krebs, H. A. and Eggleston, L. V., Ibid. 1938, 32. 913.

Krebs, H. A. and Johnson, W. A., Enzymologia 1937, 4, 148.
Krebs, H. A., Salvin, E. and Johnson, W. A., Biochem. J. 1938,
32, 113.

Krestownikoff, a., Skand. Arch. Physiol. 1927, 52, 199.

Kuyper, a. C. and Mattill, H. A., J. biol. Chem. 1933, 103, 51.
Lanford, C. Sherman, Ibid. 1939, 130, 87.

Langecker, H., Biochem. Z., 1934, 273, 43.

Lehmann, J., Botaniska Notiser, 1922, 289.

Lehmann, J., Skand. Arch. Physiol. 1938, 80, 237
Lenn6r, a., Acta obsletr. g>-n. Scand. 1934, 14, Suppl. 1-
Lindholm, H., Skand. Arch. Physiol. 1934, 70, 295.

Lundsgaabd, E., C. r. Lab. Carlsberg. 1938, 22, 333-
Lundsgaard, E., Nielsen, N. A. and 0RSKOV, S. L., Skand. Arch.
Physiol. 1936, 73, 296.

Lynen, F. and Neciullah, N., Liebigs Annul. Chemie, 1939, 541, 20o.
Mann, F. C. and Magath, T. B., Ergebn. Physiol. 1924, 23, 212.
Martius, C., Z. physiol. Chem. 1937, 247, 104: 1938, 257, 29.
Martius, C., Ergebn. EnzjTOforsch. 1939, 8, 247.

Me Michael, J., Quart. J. exp. Physiol. 1937, 27. 73.

MArtensson, j., Skand, Arch. Physiol. 1938, 80, 303; 1939, 83, 11 .
MArtensson, j., Nord. Med. 1940, 5, 253,

NordbO, R. and Scherst^N, B.. Skand. Arch. Physiol. 1931, 63, 124.
ORTEN, J. M. and SMITH, A. H., J. biol. Chem. 1937, 117, 55 d; 1938,
124, 43; 1939, 128, 101.

.Orten, j. M. and Smith, A. H., Proc. Soc. exp. Biol. N. A. 194.,

Orte^n’ f M. and Smith, A. H., Kongressbericht XVI. int. Physiol.
Kongress. Zurich 1938, H, 28/.

JOHAN mArTENSSON

PucHER, G. W., Sherman, C. C. and Vickery, H. B., J. biol. Chem.
1936, 113, 235.

Quastel, J. H. and Wooldridge, W. R., Biocheni. J. 1928, 22, 689.
^ALANT, W. and Wise, L. E., J. biol. Chem. 1916, 113, 27.

' ScnERSTEN, B., Skand. Arch. Physiol. 1931, 03, 97; 1936, 74, Suppl. 7.
Schoenheimer, R., Ratner, S. and Rittenberg, D., J. biol. Chem.

1939, 130, 703.

Sherman, C. C., Mendel, L. B. and Smith, A. H., Ibid. 1936, 113,
247; 1936, 113, 265. _

Slmola, P. E., Skand, Arch.. Physiol. 1938, 80, 375.

SiWE, S., Ibid. 1938, 80, 385.

SjOstrOm, P., Acta chir. Scand. 1937, 79, Suppl. 49.

Smith, A. H. and Meyer, C. E., J. biol. Chem. 1939, 131, 45.
S.MiTii, A. H., Orten, j. M., Johnston, C. G. and Banguess, L. C.,
Proc. Soc. exp. Biol. Med. 1939, 42, 288.

Stadie, W. C., Zapp, j. a. and Lukens, F. D. W., J. biol. Chem.

1940, 132, 411.

SvEDBERG, A., Maddock, S. and Drury, D. D., Am. J. Physiol. 1938,-
121, 209.

Thomas, J., Enzymologia 1939, 7, 231.

Thunberg, T., Skand. Arch. Physiol, 1911, 24, 23; 1911, 24, 72;

1911, 25, 37; 1936, 74, 1.

Thunberg, T., Biochem. Z. 1929, 200, 109.

Thunberg, T., Lunds Univ. Arsskrift N, F. Avd. 2. 1929, 25, Nr. 9.
Thunberg, T,, Handb. biol. Arb. Meth. 1933, Abt. IV, 2017.
Thunberg, T., Kungl, Fysiogr. Sallsk. Lund Forhandl. 1933, 3, Nr 17.
Thunberg, T., Acta path, microbiol. .scand. 1933, Suppl. 16, 535,
Wagner-Jauregg, T. and Rauen, H., Z. physiol. Chem. 1935, 237,
227.

Weiimer, C., Die Pflanzenstoffe, Jena, 1931.

Westerlund, a., Arsskrift for Lantbruks- och Mejeriinstitutel,
Alnarp, 1931.

Wishart, G. M., Biochem. J. 1923, 17, 103,

OsTBERG, O., Skand. Arch. Physiol. 1931, 02, 142; 1934, 07, 265.
OsTRERG, O., Z. gesamt. exp. Med. 1934, 94, 442.

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Acta Physiologica Scandinavica"

See other formats